init

ollama/convert/sentencepiece_model.proto

+// Copyright 2016 Google Inc.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.!
+
+syntax = "proto2";
+
+// TODO(taku): Needs to use LITE RUNTIME in OSS release.
+option optimize_for = LITE_RUNTIME;
+option go_package = "./sentencepiece";
+
+package sentencepiece;
+
+// TrainerSpec encodes a various parameters for SentencePiece training.
+// Next id: 55
+message TrainerSpec {
+  ///////////////////////////////////////////////////////////////////
+  // General parameters
+  //
+  // Input corpus files.
+  //  Trainer accepts the following two formats:
+  //  A) Monolingual: plain text, one sentence per line.
+  //  B) Bilingual:   TSV, source sentence <tab> target sentence
+  //  When bilingual data is passed, shared vocabulary model is built.
+  //  Note that the input file must be raw corpus, not a preprocessed corpus.
+  //  Trainer only loads the first `input_sentence_size` sentences specified
+  //  with this parameter.
+  repeated string input = 1;
+
+  // Input corpus format:
+  // "text": one-sentence-per-line text format (default)
+  // "tsv":  sentence <tab> freq
+  optional string input_format = 7;
+
+  // Output model file prefix.
+  // <model_prefix>.model and <model_prefix>.vocab are generated.
+  optional string model_prefix = 2;
+
+  // Model type. only have UNIGRAM now.
+  enum ModelType {
+    UNIGRAM = 1;  // Unigram language model with dynamic algorithm
+    BPE = 2;      // Byte Pair Encoding
+    WORD = 3;     // Delimitered by whitespace.
+    CHAR = 4;     // tokenizes into character sequence
+  }
+  optional ModelType model_type = 3 [default = UNIGRAM];
+
+  // Vocabulary size. 8k is the default size.
+  optional int32 vocab_size = 4 [default = 8000];
+
+  // List of the languages this model can accept.
+  // Since the model is language-agnostic, this field is used as a reference.
+  repeated string accept_language = 5;
+
+  // Size of self-test samples, which are encoded in the model file.
+  optional int32 self_test_sample_size = 6 [default = 0];
+
+  // Whether to use DP version of sentencepiece. Use it with TSV input format
+  // (requires precomputed word tab counts to work).
+  optional bool enable_differential_privacy = 50 [default = false];
+  // Set these parameters if you need DP version of sentencepiece.
+  // std of noise to add.
+  optional float differential_privacy_noise_level = 51 [default = 0.0];
+  // Clipping threshold to apply after adding noise. All the words with
+  // frequency less than this value are dropped.
+  optional uint64 differential_privacy_clipping_threshold = 52 [default = 0];
+
+  ///////////////////////////////////////////////////////////////////
+  // Training parameters.
+  //
+  // Uses characters which cover the corpus with the ratio of `chars_coverage`.
+  // This parameter determines the set of basic Alphabet of sentence piece.
+  // 1.0 - `chars_coverage` characters are treated as UNK.
+  // See also required_chars field.
+  optional float character_coverage = 10 [default = 0.9995];
+
+  // Maximum size of sentences the trainer loads from `input` parameter.
+  // Trainer simply loads the `input` files in sequence.
+  // It is better to shuffle the input corpus randomly.
+  optional uint64 input_sentence_size = 11 [default = 0];
+  optional bool shuffle_input_sentence = 19 [default = true];
+
+  // Maximum size of sentences to make seed sentence pieces.
+  // Extended suffix array is constructed to extract frequent
+  // sub-strings from the corpus. This uses 20N working space,
+  // where N is the size of corpus.
+  optional int32 mining_sentence_size = 12 [deprecated = true];
+
+  // Maximum size of sentences to train sentence pieces.
+  optional int32 training_sentence_size = 13 [deprecated = true];
+
+  // The size of seed sentencepieces.
+  // `seed_sentencepiece_size` must be larger than `vocab_size`.
+  optional int32 seed_sentencepiece_size = 14 [default = 1000000];
+
+  // In every EM sub-iterations, keeps top
+  // `shrinking_factor` * `current sentencepieces size` with respect to
+  // the loss of the sentence piece. This value should be smaller than 1.0.
+  optional float shrinking_factor = 15 [default = 0.75];
+
+  // The maximum sentence length in byte. The sentences with the length
+  // larger than `max_sentence_length` is simply ignored.
+  // Longer input tends to bring the following risks:
+  //  * Overflow during EM training (unigram language model only)
+  //  * Performance drop because of O(n log n) cost in BPE.
+  optional int32 max_sentence_length = 18 [default = 4192];
+
+  // Number of threads in the training.
+  optional int32 num_threads = 16 [default = 16];
+
+  // Number of EM sub iterations.
+  optional int32 num_sub_iterations = 17 [default = 2];
+
+  ///////////////////////////////////////////////////////////////////
+  // SentencePiece parameters which control the shapes of sentence piece.
+  //
+  // Maximum length of sentencepiece.
+  optional int32 max_sentencepiece_length = 20 [default = 16];
+
+  // Uses Unicode script to split sentence pieces.
+  // When `split_by_unicode_script` is true, we do not allow sentence piece to
+  // include multiple Unicode scripts, e.g. "F1" is not a valid piece.
+  // Exception: CJ characters (Hiragana/Katakana/Han) are all handled
+  // as one script type, since Japanese word can consist of multiple scripts.
+  // This exception is always applied regardless of the accept-language
+  // parameter.
+  optional bool split_by_unicode_script = 21 [default = true];
+
+  // When `split_by_number` is true, put a boundary between number and
+  // non-number transition. If we want to treat "F1" is one token, set this flag
+  // to be false.
+  optional bool split_by_number = 23 [default = true];
+
+  // Use a white space to split sentence pieces.
+  // When `split_by_whitespace` is false, we may have the piece containing
+  // a white space in the middle. e.g., "in_the".
+  optional bool split_by_whitespace = 22 [default = true];
+
+  // Adds whitespace symbol (_) as a suffix instead of prefix. e.g., _hello =>
+  // hello_. When `treat_whitespace_as_suffix` is true,
+  // NormalizerSpec::add_dummy_prefix will add the dummy whitespace to the end
+  // of sentence.
+  optional bool treat_whitespace_as_suffix = 24 [default = false];
+
+  // Allows pieces that only contain whitespaces instead of appearing only as
+  // prefix or suffix of other pieces.
+  optional bool allow_whitespace_only_pieces = 26 [default = false];
+
+  // Split all digits (0-9) into separate pieces.
+  optional bool split_digits = 25 [default = false];
+
+  // Defines the pre-tokenization delimiter.
+  // When specified, no pieces crossing this delimiter is not included
+  // in the vocab. Then the delimiter string is virtually ignored
+  // during the training. This field can allows constraints on the vocabulary
+  // selection. Note that this field is available on unigram mode.
+  optional string pretokenization_delimiter = 53 [ default = ""];
+
+  ///////////////////////////////////////////////////////////////////
+  // Vocabulary management
+  //
+  // Defines control symbols used as an indicator to
+  // change the behavior of the decoder. <s> and </s> are pre-defined.
+  // We can use this field to encode various meta information,
+  // including language indicator in multilingual model.
+  // These symbols are not visible to users, but visible to
+  // the decoder. Note that when the input sentence contains control symbols,
+  // they are not treated as one token, but segmented into normal pieces.
+  // Control symbols must be inserted independently from the segmentation.
+  repeated string control_symbols = 30;
+
+  // Defines user defined symbols.
+  // These symbols are added with extremely high score
+  // so they are always treated as one unique symbol in any context.
+  // Typical usage of user_defined_symbols is placeholder for named entities.
+  repeated string user_defined_symbols = 31;
+
+  // Defines required characters. Each UTF8 character in this string is included
+  // in the character set regardless of character_coverage value. Unlike
+  // user_defined_symbols, these characters have scores based on the frequency
+  // on input sentences, and the model can form subwords using characters
+  // in this field.
+  optional string required_chars = 36;
+
+  // Decomposes unknown pieces into UTF-8 bytes.
+  optional bool byte_fallback = 35 [default = false];
+
+  // When creating the vocabulary file, defines whether or not to additionally
+  // output the score for each piece.
+  optional bool vocabulary_output_piece_score = 32 [default = true];
+
+  // `vocab_size` is treated as hard limit. Crash if
+  // the model can not produce the vocab of size `vocab_size`,
+  // When `hard_vocab_limit` is false, vocab_size is treated
+  // as soft limit. Note that when model_type=char,
+  // always assumes hard_vocab_limit = false.
+  optional bool hard_vocab_limit = 33 [default = true];
+
+  // use all symbols for vocab extraction. This flag is valid
+  // if model type is either CHAR or WORD
+  optional bool use_all_vocab = 34 [default = false];
+
+  ///////////////////////////////////////////////////////////////////
+  // Reserved special meta tokens.
+  // * -1 is not used.
+  // * unk_id must not be -1.
+  // Id must starts with 0 and be contigous.
+  optional int32 unk_id = 40 [default = 0];   // <unk>
+  optional int32 bos_id = 41 [default = 1];   // <s>
+  optional int32 eos_id = 42 [default = 2];   // </s>
+  optional int32 pad_id = 43 [default = -1];  // <pad> (padding)
+  optional string unk_piece = 45 [default = "<unk>"];
+  optional string bos_piece = 46 [default = "<s>"];
+  optional string eos_piece = 47 [default = "</s>"];
+  optional string pad_piece = 48 [default = "<pad>"];
+
+  // Encodes <unk> into U+2047 (DOUBLE QUESTION MARK),
+  // since this character can be useful both for user and
+  // developer. We can easily figure out that <unk> is emitted.
+  optional string unk_surface = 44 [default = " \xE2\x81\x87 "];
+
+  // Increase bit depth to allow unigram model training on large
+  // (>10M sentences) corpora. A Side-effect of enabling this flag
+  // is increased memory usage.
+  optional bool train_extremely_large_corpus = 49 [default = false];
+
+ // Path to a seed sentencepieces file, with one tab-separated
+  // seed sentencepiece <tab> frequency per line.
+  optional string seed_sentencepieces_file = 54 [default = ""];
+
+  // Customized extensions: the range of field numbers
+  // are open to third-party extensions.
+  extensions 200 to max;
+}
+
+// NormalizerSpec encodes a various parameters for string normalizaiton
+message NormalizerSpec {
+  // name of normalization rule.
+  optional string name = 1;
+
+  // Pre-compiled normalization rule created by
+  // Builder::GetPrecompiledCharsMap() or Builder::CompileCharsMap() method.
+  // Usually this field is set by Builder::GetNormalizerSpec() method.
+  optional bytes precompiled_charsmap = 2;
+
+  // Adds dummy whitespace at the beginning of text in order to
+  // treat "world" in "world" and "hello world" in the same way.
+  optional bool add_dummy_prefix = 3 [default = true];
+
+  // Removes leading, trailing, and duplicate internal whitespace.
+  optional bool remove_extra_whitespaces = 4 [default = true];
+
+  // Replaces whitespace with meta symbol.
+  // This field must be true to train sentence piece model.
+  optional bool escape_whitespaces = 5 [default = true];
+
+  // Custom normalization rule file in TSV format.
+  // https://github.com/google/sentencepiece/blob/master/doc/normalization.md
+  // This field is only used in SentencePieceTrainer::Train() method, which
+  // compiles the rule into the binary rule stored in `precompiled_charsmap`.
+  optional string normalization_rule_tsv = 6;
+
+  // Customized extensions: the range of field numbers
+  // are open to third-party extensions.
+  extensions 200 to max;
+}
+
+// Proto to store samples for self-testing.
+message SelfTestData {
+  message Sample {
+    optional string input = 1;
+    optional string expected = 2;
+  }
+  repeated Sample samples = 1;
+
+  // Customized extensions: the range of field numbers
+  // are open to third-party extensions.
+  extensions 200 to max;
+}
+
+// ModelProto stores model parameters.
+// SentencePieceProcessor is supposed to be self-contained.
+// All settings/parameters which may change the behavior must be encoded
+// in ModelProto.
+message ModelProto {
+  message SentencePiece {
+    enum Type {
+      NORMAL = 1;        // normal symbol
+      UNKNOWN = 2;       // unknown symbol. only <unk> for now.
+      CONTROL = 3;       // control symbols. </s>, <s>, <2ja> etc.
+      USER_DEFINED = 4;  // user defined symbols.
+                         // Typical usage of USER_DEFINED symbol
+                         // is placeholder.
+      BYTE = 6;          // byte symbols. Used when `byte_fallback` is true.
+      UNUSED = 5;        // this piece is not used.
+    }
+    optional string piece = 1;  // piece must not be empty.
+    optional float score = 2;
+    optional Type type = 3 [default = NORMAL];
+
+    // Customized extensions: the range of field numbers
+    // are open to third-party extensions.
+    extensions 200 to max;
+  }
+
+  // Sentence pieces with scores.
+  repeated SentencePiece pieces = 1;
+
+  // Spec used to generate this model file.
+  optional TrainerSpec trainer_spec = 2;
+
+  // Spec for text normalization.
+  optional NormalizerSpec normalizer_spec = 3;
+
+  // Stores sample input and its expected segmentation to verify the model.
+  optional SelfTestData self_test_data = 4;
+
+  // Spec for text de-normalization.
+  optional NormalizerSpec denormalizer_spec = 5;
+
+  // Customized extensions: the range of field numbers
+  // are open to third-party extensions.
+  extensions 200 to max;
+}

ollama/convert/tokenizer.go

+package convert
+
+import (
+	"cmp"
+	"crypto/sha256"
+	"encoding/json"
+	"fmt"
+	"log/slog"
+	"os"
+	"slices"
+
+	"golang.org/x/exp/maps"
+)
+
+type Tokenizer struct {
+	Version     string         `json:"version"`
+	AddedTokens []Token        `json:"added_tokens"`
+	Model       TokenizerModel `json:"model"`
+
+	PreTokenizer struct {
+		PreTokenizers []struct {
+			Type    string `json:"type"`
+			Pattern struct {
+				Regex string `json:"Regex"`
+			} `json:"pattern"`
+		} `json:"pretokenizers"`
+	} `json:"pre_tokenizer"`
+}
+
+type TokenizerModel struct {
+	Type   string         `json:"type"`
+	Vocab  map[string]int `json:"vocab"`
+	Merges []string       `json:"merges"`
+	Tokens []Token
+}
+
+type Token struct {
+	ID          int    `json:"id"`
+	Content     string `json:"content"`
+	Special     bool   `json:"special"`
+	UserDefined bool
+}
+
+func (t *Token) Type() int32 {
+	switch {
+	case t.Special:
+		return tokenTypeControl
+	case t.UserDefined:
+		return tokenTypeUserDefined
+	default:
+		return tokenTypeNormal
+	}
+}
+
+func (t *Tokenizer) maxID() int {
+	return max(
+		slices.Max(maps.Values(t.Model.Vocab)),
+		slices.MaxFunc(t.AddedTokens, func(a, b Token) int {
+			return cmp.Compare(a.ID, b.ID)
+		}).ID,
+	)
+}
+
+func parseTokens(dirpath string) (pre string, tokens []Token, merges []string, err error) {
+	f, err := os.Open(dirpath)
+	if err != nil {
+		panic(err)
+	}
+	defer f.Close()
+
+	var t Tokenizer
+	if err := json.NewDecoder(f).Decode(&t); err != nil {
+		return "", nil, nil, err
+	}
+
+	tokens = make([]Token, t.maxID()+1)
+	for k, v := range t.Model.Vocab {
+		tokens[v] = Token{ID: v, Content: k, Special: false, UserDefined: false}
+	}
+
+	for _, v := range t.AddedTokens {
+		v.UserDefined = true
+		tokens[v.ID] = v
+	}
+
+	sha256sum := sha256.New()
+	for _, pt := range t.PreTokenizer.PreTokenizers {
+		if pt.Type == "Split" && pt.Pattern.Regex != "" {
+			sha256sum.Write([]byte(pt.Pattern.Regex))
+		}
+	}
+
+	switch digest := fmt.Sprintf("%x", sha256sum.Sum(nil)); digest {
+	case "d98f9631be1e9607a9848c26c1f9eac1aa9fc21ac6ba82a2fc0741af9780a48f":
+		pre = "llama-bpe"
+	case "03df5c5863ad70781dcfdef491ead25140f895fe8010964be0daefe27be32b02":
+		pre = "deepseek-llm"
+	case "21cde974d587f0d54dc8d56b183cc1e6239600172035c68fbd6d4b9f8da0576e":
+		pre = "deepseek-coder"
+	default:
+		slog.Warn("unknown pretokenizer, using default", "digest", digest)
+		pre = "default"
+	}
+
+	return pre, tokens, t.Model.Merges, nil
+}

ollama/convert/torch.go

+package convert
+
+import (
+	"encoding/binary"
+	"encoding/json"
+	"fmt"
+	"io"
+	"log/slog"
+	"os"
+	"path/filepath"
+	"regexp"
+	"strings"
+
+	"github.com/nlpodyssey/gopickle/pytorch"
+	"github.com/nlpodyssey/gopickle/types"
+	"github.com/x448/float16"
+
+	"github.com/ollama/ollama/llm"
+)
+
+type torchWriterTo struct {
+	t *llm.Tensor
+
+	params *Params
+	bo     ByteOrder
+
+	storage  pytorch.StorageInterface
+	repacker func(string, []float32, []uint64) ([]float32, error)
+}
+
+type TorchFormat struct{}
+
+func (tf *TorchFormat) GetTensors(dirpath string, params *Params) ([]llm.Tensor, error) {
+	slog.Debug("getting torch tensors")
+
+	var files []string
+	if pt, _ := filepath.Glob(filepath.Join(dirpath, "consolidated*.pth")); len(pt) > 0 {
+		files = append(files, pt...)
+	} else if pt, _ := filepath.Glob(filepath.Join(dirpath, "pytorch_model*.pth")); len(pt) > 0 {
+		files = append(files, pt...)
+	}
+
+	var offset uint64
+	var tensors []llm.Tensor
+	for _, fn := range files {
+		m, err := pytorch.Load(fn)
+		if err != nil {
+			slog.Error(fmt.Sprintf("error unpickling: %q", err))
+			return []llm.Tensor{}, err
+		}
+
+		for _, k := range m.(*types.Dict).Keys() {
+			if strings.HasSuffix(k.(string), "self_attn.rotary_emb.inv_freq") {
+				continue
+			}
+
+			t, _ := m.(*types.Dict).Get(k)
+			tshape := t.(*pytorch.Tensor).Size
+
+			var size uint64
+			var kind uint32
+			switch len(tshape) {
+			case 0:
+				continue
+			case 1:
+				// convert to float32
+				kind = 0
+				size = uint64(tshape[0] * 4)
+			case 2:
+				// convert to float16
+				kind = 1
+				size = uint64(tshape[0] * tshape[1] * 2)
+			}
+
+			ggufName, err := tf.GetLayerName(k.(string))
+			if err != nil {
+				slog.Error(err.Error())
+				return nil, err
+			}
+			slog.Debug(fmt.Sprintf("'%35s': '%30s' %10d [%#v]", k.(string), ggufName, size, tshape))
+
+			shape := []uint64{0, 0, 0, 0}
+			for i := range tshape {
+				shape[i] = uint64(tshape[i])
+			}
+
+			tensor := llm.Tensor{
+				Name:   ggufName,
+				Kind:   kind,
+				Offset: offset, // calculate the offset
+				Shape:  shape,
+			}
+
+			tensor.WriterTo = torchWriterTo{
+				t:       &tensor,
+				params:  params,
+				bo:      params.ByteOrder,
+				storage: t.(*pytorch.Tensor).Source,
+			}
+
+			tensors = append(tensors, tensor)
+			offset += size
+		}
+	}
+
+	return tensors, nil
+}
+
+func getAltParams(dirpath string) (*Params, error) {
+	f, err := os.Open(filepath.Join(dirpath, "params.json"))
+	if err != nil {
+		slog.Error("no params.json")
+		return nil, err
+	}
+	defer f.Close()
+
+	type TorchParams struct {
+		HiddenSize     int     `json:"dim"`
+		AttentionHeads int     `json:"n_heads"`
+		KeyValHeads    int     `json:"n_kv_heads"`
+		HiddenLayers   int     `json:"n_layers"`
+		RopeTheta      float64 `json:"rope_theta"`
+		NormEPS        float64 `json:"norm_eps"`
+	}
+
+	var tparams TorchParams
+
+	d := json.NewDecoder(f)
+	err = d.Decode(&tparams)
+	if err != nil {
+		return nil, err
+	}
+
+	params := &Params{
+		Architectures:  []string{"LlamaForCausalLM"},
+		HiddenSize:     tparams.HiddenSize,
+		AttentionHeads: tparams.AttentionHeads,
+		KeyValHeads:    tparams.KeyValHeads,
+		HiddenLayers:   tparams.HiddenLayers,
+		NormEPS:        tparams.NormEPS,
+	}
+
+	switch {
+	case tparams.RopeTheta == 1000000:
+		// Codellama
+		params.ContextSize = 16384
+	case tparams.NormEPS == 1e-06:
+		// llama2
+		slog.Debug("Found llama2 - setting context size to 4096")
+		params.ContextSize = 4096
+	default:
+		params.ContextSize = 2048
+	}
+
+	params.ByteOrder = binary.LittleEndian
+	return params, nil
+}
+
+func (m *TorchFormat) GetParams(dirpath string) (*Params, error) {
+	f, err := os.Open(filepath.Join(dirpath, "config.json"))
+	if err != nil {
+		if os.IsNotExist(err) {
+			// try params.json instead
+			return getAltParams(dirpath)
+		} else {
+			return nil, err
+		}
+	}
+
+	var params Params
+	d := json.NewDecoder(f)
+	err = d.Decode(&params)
+	if err != nil {
+		return nil, err
+	}
+
+	params.ByteOrder = binary.LittleEndian
+	return &params, nil
+}
+
+func (m *TorchFormat) GetLayerName(n string) (string, error) {
+	directMap := map[string]string{
+		"tok_embeddings.weight":     "token_embd.weight",
+		"output.weight":             "output.weight",
+		"norm.weight":               "output_norm.weight",
+		"rope.freqs":                "rope_freqs.weight",
+		"model.embed_tokens.weight": "token_embd.weight",
+		"lm_head.weight":            "output.weight",
+		"model.norm.weight":         "output_norm.weight",
+	}
+
+	lMap := map[string]string{
+		"layers.(\\d+).attention_norm.weight":                 "blk.$1.attn_norm.weight",
+		"layers.(\\d+).attention_output_norm.weight":          "blk.$1.attn_norm.weight",
+		"layers.(\\d+).feed_forward.w2.weight":                "blk.$1.ffn_down.weight",
+		"layers.(\\d+).feed_forward.w1.weight":                "blk.$1.ffn_gate.weight",
+		"layers.(\\d+).feed_forward.w3.weight":                "blk.$1.ffn_up.weight",
+		"layers.(\\d+).ffn_norm.weight":                       "blk.$1.ffn_norm.weight",
+		"layers.(\\d+).attention.wk.weight":                   "blk.$1.attn_k.weight",
+		"layers.(\\d+).attention.wo.weight":                   "blk.$1.attn_output.weight",
+		"layers.(\\d+).attention.wq.weight":                   "blk.$1.attn_q.weight",
+		"layers.(\\d+).attention.wv.weight":                   "blk.$1.attn_v.weight",
+		"model.layers.(\\d+).input_layernorm.weight":          "blk.$1.attn_norm.weight",
+		"model.layers.(\\d+).mlp.down_proj.weight":            "blk.$1.ffn_down.weight",
+		"model.layers.(\\d+).mlp.gate_proj.weight":            "blk.$1.ffn_gate.weight",
+		"model.layers.(\\d+).mlp.up_proj.weight":              "blk.$1.ffn_up.weight",
+		"model.layers.(\\d+).post_attention_layernorm.weight": "blk.$1.ffn_norm.weight",
+		"model.layers.(\\d+).self_attn.k_proj.weight":         "blk.$1.attn_k.weight",
+		"model.layers.(\\d+).self_attn.o_proj.weight":         "blk.$1.attn_output.weight",
+		"model.layers.(\\d+).self_attn.q_proj.weight":         "blk.$1.attn_q.weight",
+		"model.layers.(\\d+).self_attn.v_proj.weight":         "blk.$1.attn_v.weight",
+	}
+
+	v, ok := directMap[n]
+	if ok {
+		return v, nil
+	}
+
+	// quick hack to rename the layers to gguf format
+	for k, v := range lMap {
+		re := regexp.MustCompile(k)
+		newName := re.ReplaceAllString(n, v)
+		if newName != n {
+			return newName, nil
+		}
+	}
+
+	return "", fmt.Errorf("couldn't find a layer name for '%s'", n)
+}
+
+func (r torchWriterTo) WriteTo(w io.Writer) (n int64, err error) {
+	var f32s []float32
+	switch s := r.storage.(type) {
+	case *pytorch.FloatStorage:
+		f32s = s.Data
+	case *pytorch.HalfStorage:
+		f32s = s.Data
+	case *pytorch.BFloat16Storage:
+		f32s = s.Data
+	default:
+		return 0, fmt.Errorf("unknown data type: %T", s)
+	}
+
+	if r.repacker != nil {
+		f32s, err = r.repacker(r.t.Name, f32s, r.t.Shape)
+		if err != nil {
+			return 0, err
+		}
+	}
+
+	switch r.t.Kind {
+	case 0:
+		return 0, binary.Write(w, r.bo, f32s)
+	case 1:
+		f16s := make([]uint16, len(f32s))
+		for i := range f32s {
+			f16s[i] = float16.Fromfloat32(f32s[i]).Bits()
+		}
+
+		return 0, binary.Write(w, r.bo, f16s)
+	default:
+		return 0, fmt.Errorf("unknown storage type: %d", r.t.Kind)
+	}
+}
+
+func (m *TorchFormat) GetModelArch(name, dirPath string, params *Params) (ModelArch, error) {
+	switch len(params.Architectures) {
+	case 0:
+		return nil, fmt.Errorf("No architecture specified to convert")
+	case 1:
+		switch params.Architectures[0] {
+		case "LlamaForCausalLM":
+			return &LlamaModel{
+				ModelData{
+					Name:   name,
+					Path:   dirPath,
+					Params: params,
+					Format: m,
+				},
+			}, nil
+		default:
+			return nil, fmt.Errorf("Models based on '%s' are not yet supported", params.Architectures[0])
+		}
+	}
+
+	return nil, fmt.Errorf("Unknown error")
+}

ollama/docs/README.md

+# Documentation
+
+### Getting Started
+* [Quickstart](../README.md#quickstart)
+* [Examples](../examples)
+* [Importing models](./import.md)
+* [Linux Documentation](./linux.md)
+* [Windows Documentation](./windows.md)
+* [Docker Documentation](./docker.md)
+
+### Reference
+
+* [API Reference](./api.md)
+* [Modelfile Reference](./modelfile.md)
+* [OpenAI Compatibility](./openai.md)
+
+### Resources
+
+* [Troubleshooting Guide](./troubleshooting.md)
+* [FAQ](./faq.md)
+* [Development guide](./development.md)

ollama/docs/api.md

+# API
+
+## Endpoints
+
+- [Generate a completion](#generate-a-completion)
+- [Generate a chat completion](#generate-a-chat-completion)
+- [Create a Model](#create-a-model)
+- [List Local Models](#list-local-models)
+- [Show Model Information](#show-model-information)
+- [Copy a Model](#copy-a-model)
+- [Delete a Model](#delete-a-model)
+- [Pull a Model](#pull-a-model)
+- [Push a Model](#push-a-model)
+- [Generate Embeddings](#generate-embeddings)
+- [List Running Models](#list-running-models)
+
+## Conventions
+
+### Model names
+
+Model names follow a `model:tag` format, where `model` can have an optional namespace such as `example/model`. Some examples are `orca-mini:3b-q4_1` and `llama3:70b`. The tag is optional and, if not provided, will default to `latest`. The tag is used to identify a specific version.
+
+### Durations
+
+All durations are returned in nanoseconds.
+
+### Streaming responses
+
+Certain endpoints stream responses as JSON objects and can optional return non-streamed responses.
+
+## Generate a completion
+
+```shell
+POST /api/generate
+```
+
+Generate a response for a given prompt with a provided model. This is a streaming endpoint, so there will be a series of responses. The final response object will include statistics and additional data from the request.
+
+### Parameters
+
+- `model`: (required) the [model name](#model-names)
+- `prompt`: the prompt to generate a response for
+- `images`: (optional) a list of base64-encoded images (for multimodal models such as `llava`)
+
+Advanced parameters (optional):
+
+- `format`: the format to return a response in. Currently the only accepted value is `json`
+- `options`: additional model parameters listed in the documentation for the [Modelfile](./modelfile.md#valid-parameters-and-values) such as `temperature`
+- `system`: system message to (overrides what is defined in the `Modelfile`)
+- `template`: the prompt template to use (overrides what is defined in the `Modelfile`)
+- `context`: the context parameter returned from a previous request to `/generate`, this can be used to keep a short conversational memory
+- `stream`: if `false` the response will be returned as a single response object, rather than a stream of objects
+- `raw`: if `true` no formatting will be applied to the prompt. You may choose to use the `raw` parameter if you are specifying a full templated prompt in your request to the API
+- `keep_alive`: controls how long the model will stay loaded into memory following the request (default: `5m`)
+
+#### JSON mode
+
+Enable JSON mode by setting the `format` parameter to `json`. This will structure the response as a valid JSON object. See the JSON mode [example](#request-json-mode) below.
+
+> Note: it's important to instruct the model to use JSON in the `prompt`. Otherwise, the model may generate large amounts whitespace.
+
+### Examples
+
+#### Generate request (Streaming)
+
+##### Request
+
+```shell
+curl http://localhost:11434/api/generate -d '{
+  "model": "llama3",
+  "prompt": "Why is the sky blue?"
+}'
+```
+
+##### Response
+
+A stream of JSON objects is returned:
+
+```json
+{
+  "model": "llama3",
+  "created_at": "2023-08-04T08:52:19.385406455-07:00",
+  "response": "The",
+  "done": false
+}
+```
+
+The final response in the stream also includes additional data about the generation:
+
+- `total_duration`: time spent generating the response
+- `load_duration`: time spent in nanoseconds loading the model
+- `prompt_eval_count`: number of tokens in the prompt
+- `prompt_eval_duration`: time spent in nanoseconds evaluating the prompt
+- `eval_count`: number of tokens in the response
+- `eval_duration`: time in nanoseconds spent generating the response
+- `context`: an encoding of the conversation used in this response, this can be sent in the next request to keep a conversational memory
+- `response`: empty if the response was streamed, if not streamed, this will contain the full response
+
+To calculate how fast the response is generated in tokens per second (token/s), divide `eval_count` / `eval_duration` * `10^9`.
+
+```json
+{
+  "model": "llama3",
+  "created_at": "2023-08-04T19:22:45.499127Z",
+  "response": "",
+  "done": true,
+  "context": [1, 2, 3],
+  "total_duration": 10706818083,
+  "load_duration": 6338219291,
+  "prompt_eval_count": 26,
+  "prompt_eval_duration": 130079000,
+  "eval_count": 259,
+  "eval_duration": 4232710000
+}
+```
+
+#### Request (No streaming)
+
+##### Request
+
+A response can be received in one reply when streaming is off.
+
+```shell
+curl http://localhost:11434/api/generate -d '{
+  "model": "llama3",
+  "prompt": "Why is the sky blue?",
+  "stream": false
+}'
+```
+
+##### Response
+
+If `stream` is set to `false`, the response will be a single JSON object:
+
+```json
+{
+  "model": "llama3",
+  "created_at": "2023-08-04T19:22:45.499127Z",
+  "response": "The sky is blue because it is the color of the sky.",
+  "done": true,
+  "context": [1, 2, 3],
+  "total_duration": 5043500667,
+  "load_duration": 5025959,
+  "prompt_eval_count": 26,
+  "prompt_eval_duration": 325953000,
+  "eval_count": 290,
+  "eval_duration": 4709213000
+}
+```
+
+#### Request (JSON mode)
+
+> When `format` is set to `json`, the output will always be a well-formed JSON object. It's important to also instruct the model to respond in JSON.
+
+##### Request
+
+```shell
+curl http://localhost:11434/api/generate -d '{
+  "model": "llama3",
+  "prompt": "What color is the sky at different times of the day? Respond using JSON",
+  "format": "json",
+  "stream": false
+}'
+```
+
+##### Response
+
+```json
+{
+  "model": "llama3",
+  "created_at": "2023-11-09T21:07:55.186497Z",
+  "response": "{\n\"morning\": {\n\"color\": \"blue\"\n},\n\"noon\": {\n\"color\": \"blue-gray\"\n},\n\"afternoon\": {\n\"color\": \"warm gray\"\n},\n\"evening\": {\n\"color\": \"orange\"\n}\n}\n",
+  "done": true,
+  "context": [1, 2, 3],
+  "total_duration": 4648158584,
+  "load_duration": 4071084,
+  "prompt_eval_count": 36,
+  "prompt_eval_duration": 439038000,
+  "eval_count": 180,
+  "eval_duration": 4196918000
+}
+```
+
+The value of `response` will be a string containing JSON similar to:
+
+```json
+{
+  "morning": {
+    "color": "blue"
+  },
+  "noon": {
+    "color": "blue-gray"
+  },
+  "afternoon": {
+    "color": "warm gray"
+  },
+  "evening": {
+    "color": "orange"
+  }
+}
+```
+
+#### Request (with images)
+
+To submit images to multimodal models such as `llava` or `bakllava`, provide a list of base64-encoded `images`:
+
+#### Request
+
+```shell
+curl http://localhost:11434/api/generate -d '{
+  "model": "llava",
+  "prompt":"What is in this picture?",
+  "stream": false,
+  "images": ["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"]
+}'
+```
+
+#### Response
+
+```
+{
+  "model": "llava",
+  "created_at": "2023-11-03T15:36:02.583064Z",
+  "response": "A happy cartoon character, which is cute and cheerful.",
+  "done": true,
+  "context": [1, 2, 3],
+  "total_duration": 2938432250,
+  "load_duration": 2559292,
+  "prompt_eval_count": 1,
+  "prompt_eval_duration": 2195557000,
+  "eval_count": 44,
+  "eval_duration": 736432000
+}
+```
+
+#### Request (Raw Mode)
+
+In some cases, you may wish to bypass the templating system and provide a full prompt. In this case, you can use the `raw` parameter to disable templating. Also note that raw mode will not return a context.
+
+##### Request
+
+```shell
+curl http://localhost:11434/api/generate -d '{
+  "model": "mistral",
+  "prompt": "[INST] why is the sky blue? [/INST]",
+  "raw": true,
+  "stream": false
+}'
+```
+
+#### Request (Reproducible outputs)
+
+For reproducible outputs, set `seed` to a number:
+
+##### Request
+
+```shell
+curl http://localhost:11434/api/generate -d '{
+  "model": "mistral",
+  "prompt": "Why is the sky blue?",
+  "options": {
+    "seed": 123
+  }
+}'
+```
+
+##### Response
+
+```json
+{
+  "model": "mistral",
+  "created_at": "2023-11-03T15:36:02.583064Z",
+  "response": " The sky appears blue because of a phenomenon called Rayleigh scattering.",
+  "done": true,
+  "total_duration": 8493852375,
+  "load_duration": 6589624375,
+  "prompt_eval_count": 14,
+  "prompt_eval_duration": 119039000,
+  "eval_count": 110,
+  "eval_duration": 1779061000
+}
+```
+
+#### Generate request (With options)
+
+If you want to set custom options for the model at runtime rather than in the Modelfile, you can do so with the `options` parameter. This example sets every available option, but you can set any of them individually and omit the ones you do not want to override.
+
+##### Request
+
+```shell
+curl http://localhost:11434/api/generate -d '{
+  "model": "llama3",
+  "prompt": "Why is the sky blue?",
+  "stream": false,
+  "options": {
+    "num_keep": 5,
+    "seed": 42,
+    "num_predict": 100,
+    "top_k": 20,
+    "top_p": 0.9,
+    "tfs_z": 0.5,
+    "typical_p": 0.7,
+    "repeat_last_n": 33,
+    "temperature": 0.8,
+    "repeat_penalty": 1.2,
+    "presence_penalty": 1.5,
+    "frequency_penalty": 1.0,
+    "mirostat": 1,
+    "mirostat_tau": 0.8,
+    "mirostat_eta": 0.6,
+    "penalize_newline": true,
+    "stop": ["\n", "user:"],
+    "numa": false,
+    "num_ctx": 1024,
+    "num_batch": 2,
+    "num_gpu": 1,
+    "main_gpu": 0,
+    "low_vram": false,
+    "f16_kv": true,
+    "vocab_only": false,
+    "use_mmap": true,
+    "use_mlock": false,
+    "num_thread": 8
+  }
+}'
+```
+
+##### Response
+
+```json
+{
+  "model": "llama3",
+  "created_at": "2023-08-04T19:22:45.499127Z",
+  "response": "The sky is blue because it is the color of the sky.",
+  "done": true,
+  "context": [1, 2, 3],
+  "total_duration": 4935886791,
+  "load_duration": 534986708,
+  "prompt_eval_count": 26,
+  "prompt_eval_duration": 107345000,
+  "eval_count": 237,
+  "eval_duration": 4289432000
+}
+```
+
+#### Load a model
+
+If an empty prompt is provided, the model will be loaded into memory.
+
+##### Request
+
+```shell
+curl http://localhost:11434/api/generate -d '{
+  "model": "llama3"
+}'
+```
+
+##### Response
+
+A single JSON object is returned:
+
+```json
+{
+  "model": "llama3",
+  "created_at": "2023-12-18T19:52:07.071755Z",
+  "response": "",
+  "done": true
+}
+```
+
+## Generate a chat completion
+
+```shell
+POST /api/chat
+```
+
+Generate the next message in a chat with a provided model. This is a streaming endpoint, so there will be a series of responses. Streaming can be disabled using `"stream": false`. The final response object will include statistics and additional data from the request.
+
+### Parameters
+
+- `model`: (required) the [model name](#model-names)
+- `messages`: the messages of the chat, this can be used to keep a chat memory
+
+The `message` object has the following fields:
+
+- `role`: the role of the message, either `system`, `user` or `assistant`
+- `content`: the content of the message
+- `images` (optional): a list of images to include in the message (for multimodal models such as `llava`)
+
+Advanced parameters (optional):
+
+- `format`: the format to return a response in. Currently the only accepted value is `json`
+- `options`: additional model parameters listed in the documentation for the [Modelfile](./modelfile.md#valid-parameters-and-values) such as `temperature`
+- `stream`: if `false` the response will be returned as a single response object, rather than a stream of objects
+- `keep_alive`: controls how long the model will stay loaded into memory following the request (default: `5m`)
+
+### Examples
+
+#### Chat Request (Streaming)
+
+##### Request
+
+Send a chat message with a streaming response.
+
+```shell
+curl http://localhost:11434/api/chat -d '{
+  "model": "llama3",
+  "messages": [
+    {
+      "role": "user",
+      "content": "why is the sky blue?"
+    }
+  ]
+}'
+```
+
+##### Response
+
+A stream of JSON objects is returned:
+
+```json
+{
+  "model": "llama3",
+  "created_at": "2023-08-04T08:52:19.385406455-07:00",
+  "message": {
+    "role": "assistant",
+    "content": "The",
+    "images": null
+  },
+  "done": false
+}
+```
+
+Final response:
+
+```json
+{
+  "model": "llama3",
+  "created_at": "2023-08-04T19:22:45.499127Z",
+  "done": true,
+  "total_duration": 4883583458,
+  "load_duration": 1334875,
+  "prompt_eval_count": 26,
+  "prompt_eval_duration": 342546000,
+  "eval_count": 282,
+  "eval_duration": 4535599000
+}
+```
+
+#### Chat request (No streaming)
+
+##### Request
+
+```shell
+curl http://localhost:11434/api/chat -d '{
+  "model": "llama3",
+  "messages": [
+    {
+      "role": "user",
+      "content": "why is the sky blue?"
+    }
+  ],
+  "stream": false
+}'
+```
+
+##### Response
+
+```json
+{
+  "model": "registry.ollama.ai/library/llama3:latest",
+  "created_at": "2023-12-12T14:13:43.416799Z",
+  "message": {
+    "role": "assistant",
+    "content": "Hello! How are you today?"
+  },
+  "done": true,
+  "total_duration": 5191566416,
+  "load_duration": 2154458,
+  "prompt_eval_count": 26,
+  "prompt_eval_duration": 383809000,
+  "eval_count": 298,
+  "eval_duration": 4799921000
+}
+```
+
+#### Chat request (With History)
+
+Send a chat message with a conversation history. You can use this same approach to start the conversation using multi-shot or chain-of-thought prompting.
+
+##### Request
+
+```shell
+curl http://localhost:11434/api/chat -d '{
+  "model": "llama3",
+  "messages": [
+    {
+      "role": "user",
+      "content": "why is the sky blue?"
+    },
+    {
+      "role": "assistant",
+      "content": "due to rayleigh scattering."
+    },
+    {
+      "role": "user",
+      "content": "how is that different than mie scattering?"
+    }
+  ]
+}'
+```
+
+##### Response
+
+A stream of JSON objects is returned:
+
+```json
+{
+  "model": "llama3",
+  "created_at": "2023-08-04T08:52:19.385406455-07:00",
+  "message": {
+    "role": "assistant",
+    "content": "The"
+  },
+  "done": false
+}
+```
+
+Final response:
+
+```json
+{
+  "model": "llama3",
+  "created_at": "2023-08-04T19:22:45.499127Z",
+  "done": true,
+  "total_duration": 8113331500,
+  "load_duration": 6396458,
+  "prompt_eval_count": 61,
+  "prompt_eval_duration": 398801000,
+  "eval_count": 468,
+  "eval_duration": 7701267000
+}
+```
+
+#### Chat request (with images)
+
+##### Request
+
+Send a chat message with a conversation history.
+
+```shell
+curl http://localhost:11434/api/chat -d '{
+  "model": "llava",
+  "messages": [
+    {
+      "role": "user",
+      "content": "what is in this image?",
+      "images": ["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"]
+    }
+  ]
+}'
+```
+
+##### Response
+
+```json
+{
+  "model": "llava",
+  "created_at": "2023-12-13T22:42:50.203334Z",
+  "message": {
+    "role": "assistant",
+    "content": " The image features a cute, little pig with an angry facial expression. It's wearing a heart on its shirt and is waving in the air. This scene appears to be part of a drawing or sketching project.",
+    "images": null
+  },
+  "done": true,
+  "total_duration": 1668506709,
+  "load_duration": 1986209,
+  "prompt_eval_count": 26,
+  "prompt_eval_duration": 359682000,
+  "eval_count": 83,
+  "eval_duration": 1303285000
+}
+```
+
+#### Chat request (Reproducible outputs)
+
+##### Request
+
+```shell
+curl http://localhost:11434/api/chat -d '{
+  "model": "llama3",
+  "messages": [
+    {
+      "role": "user",
+      "content": "Hello!"
+    }
+  ],
+  "options": {
+    "seed": 101,
+    "temperature": 0
+  }
+}'
+```
+
+##### Response
+
+```json
+{
+  "model": "registry.ollama.ai/library/llama3:latest",
+  "created_at": "2023-12-12T14:13:43.416799Z",
+  "message": {
+    "role": "assistant",
+    "content": "Hello! How are you today?"
+  },
+  "done": true,
+  "total_duration": 5191566416,
+  "load_duration": 2154458,
+  "prompt_eval_count": 26,
+  "prompt_eval_duration": 383809000,
+  "eval_count": 298,
+  "eval_duration": 4799921000
+}
+```
+
+## Create a Model
+
+```shell
+POST /api/create
+```
+
+Create a model from a [`Modelfile`](./modelfile.md). It is recommended to set `modelfile` to the content of the Modelfile rather than just set `path`. This is a requirement for remote create. Remote model creation must also create any file blobs, fields such as `FROM` and `ADAPTER`, explicitly with the server using [Create a Blob](#create-a-blob) and the value to the path indicated in the response.
+
+### Parameters
+
+- `name`: name of the model to create
+- `modelfile` (optional): contents of the Modelfile
+- `stream`: (optional) if `false` the response will be returned as a single response object, rather than a stream of objects
+- `path` (optional): path to the Modelfile
+
+### Examples
+
+#### Create a new model
+
+Create a new model from a `Modelfile`.
+
+##### Request
+
+```shell
+curl http://localhost:11434/api/create -d '{
+  "name": "mario",
+  "modelfile": "FROM llama3\nSYSTEM You are mario from Super Mario Bros."
+}'
+```
+
+##### Response
+
+A stream of JSON objects. Notice that the final JSON object shows a `"status": "success"`.
+
+```json
+{"status":"reading model metadata"}
+{"status":"creating system layer"}
+{"status":"using already created layer sha256:22f7f8ef5f4c791c1b03d7eb414399294764d7cc82c7e94aa81a1feb80a983a2"}
+{"status":"using already created layer sha256:8c17c2ebb0ea011be9981cc3922db8ca8fa61e828c5d3f44cb6ae342bf80460b"}
+{"status":"using already created layer sha256:7c23fb36d80141c4ab8cdbb61ee4790102ebd2bf7aeff414453177d4f2110e5d"}
+{"status":"using already created layer sha256:2e0493f67d0c8c9c68a8aeacdf6a38a2151cb3c4c1d42accf296e19810527988"}
+{"status":"using already created layer sha256:2759286baa875dc22de5394b4a925701b1896a7e3f8e53275c36f75a877a82c9"}
+{"status":"writing layer sha256:df30045fe90f0d750db82a058109cecd6d4de9c90a3d75b19c09e5f64580bb42"}
+{"status":"writing layer sha256:f18a68eb09bf925bb1b669490407c1b1251c5db98dc4d3d81f3088498ea55690"}
+{"status":"writing manifest"}
+{"status":"success"}
+```
+
+### Check if a Blob Exists
+
+```shell
+HEAD /api/blobs/:digest
+```
+
+Ensures that the file blob used for a FROM or ADAPTER field exists on the server. This is checking your Ollama server and not Ollama.ai.
+
+#### Query Parameters
+
+- `digest`: the SHA256 digest of the blob
+
+#### Examples
+
+##### Request
+
+```shell
+curl -I http://localhost:11434/api/blobs/sha256:29fdb92e57cf0827ded04ae6461b5931d01fa595843f55d36f5b275a52087dd2
+```
+
+##### Response
+
+Return 200 OK if the blob exists, 404 Not Found if it does not.
+
+### Create a Blob
+
+```shell
+POST /api/blobs/:digest
+```
+
+Create a blob from a file on the server. Returns the server file path.
+
+#### Query Parameters
+
+- `digest`: the expected SHA256 digest of the file
+
+#### Examples
+
+##### Request
+
+```shell
+curl -T model.bin -X POST http://localhost:11434/api/blobs/sha256:29fdb92e57cf0827ded04ae6461b5931d01fa595843f55d36f5b275a52087dd2
+```
+
+##### Response
+
+Return 201 Created if the blob was successfully created, 400 Bad Request if the digest used is not expected.
+
+## List Local Models
+
+```shell
+GET /api/tags
+```
+
+List models that are available locally.
+
+### Examples
+
+#### Request
+
+```shell
+curl http://localhost:11434/api/tags
+```
+
+#### Response
+
+A single JSON object will be returned.
+
+```json
+{
+  "models": [
+    {
+      "name": "codellama:13b",
+      "modified_at": "2023-11-04T14:56:49.277302595-07:00",
+      "size": 7365960935,
+      "digest": "9f438cb9cd581fc025612d27f7c1a6669ff83a8bb0ed86c94fcf4c5440555697",
+      "details": {
+        "format": "gguf",
+        "family": "llama",
+        "families": null,
+        "parameter_size": "13B",
+        "quantization_level": "Q4_0"
+      }
+    },
+    {
+      "name": "llama3:latest",
+      "modified_at": "2023-12-07T09:32:18.757212583-08:00",
+      "size": 3825819519,
+      "digest": "fe938a131f40e6f6d40083c9f0f430a515233eb2edaa6d72eb85c50d64f2300e",
+      "details": {
+        "format": "gguf",
+        "family": "llama",
+        "families": null,
+        "parameter_size": "7B",
+        "quantization_level": "Q4_0"
+      }
+    }
+  ]
+}
+```
+
+## Show Model Information
+
+```shell
+POST /api/show
+```
+
+Show information about a model including details, modelfile, template, parameters, license, and system prompt.
+
+### Parameters
+
+- `name`: name of the model to show
+
+### Examples
+
+#### Request
+
+```shell
+curl http://localhost:11434/api/show -d '{
+  "name": "llama3"
+}'
+```
+
+#### Response
+
+```json
+{
+  "modelfile": "# Modelfile generated by \"ollama show\"\n# To build a new Modelfile based on this one, replace the FROM line with:\n# FROM llava:latest\n\nFROM /Users/matt/.ollama/models/blobs/sha256:200765e1283640ffbd013184bf496e261032fa75b99498a9613be4e94d63ad52\nTEMPLATE \"\"\"{{ .System }}\nUSER: {{ .Prompt }}\nASSISTANT: \"\"\"\nPARAMETER num_ctx 4096\nPARAMETER stop \"\u003c/s\u003e\"\nPARAMETER stop \"USER:\"\nPARAMETER stop \"ASSISTANT:\"",
+  "parameters": "num_ctx                        4096\nstop                           \u003c/s\u003e\nstop                           USER:\nstop                           ASSISTANT:",
+  "template": "{{ .System }}\nUSER: {{ .Prompt }}\nASSISTANT: ",
+  "details": {
+    "format": "gguf",
+    "family": "llama",
+    "families": ["llama", "clip"],
+    "parameter_size": "7B",
+    "quantization_level": "Q4_0"
+  }
+}
+```
+
+## Copy a Model
+
+```shell
+POST /api/copy
+```
+
+Copy a model. Creates a model with another name from an existing model.
+
+### Examples
+
+#### Request
+
+```shell
+curl http://localhost:11434/api/copy -d '{
+  "source": "llama3",
+  "destination": "llama3-backup"
+}'
+```
+
+#### Response
+
+Returns a 200 OK if successful, or a 404 Not Found if the source model doesn't exist.
+
+## Delete a Model
+
+```shell
+DELETE /api/delete
+```
+
+Delete a model and its data.
+
+### Parameters
+
+- `name`: model name to delete
+
+### Examples
+
+#### Request
+
+```shell
+curl -X DELETE http://localhost:11434/api/delete -d '{
+  "name": "llama3:13b"
+}'
+```
+
+#### Response
+
+Returns a 200 OK if successful, 404 Not Found if the model to be deleted doesn't exist.
+
+## Pull a Model
+
+```shell
+POST /api/pull
+```
+
+Download a model from the ollama library. Cancelled pulls are resumed from where they left off, and multiple calls will share the same download progress.
+
+### Parameters
+
+- `name`: name of the model to pull
+- `insecure`: (optional) allow insecure connections to the library. Only use this if you are pulling from your own library during development.
+- `stream`: (optional) if `false` the response will be returned as a single response object, rather than a stream of objects
+
+### Examples
+
+#### Request
+
+```shell
+curl http://localhost:11434/api/pull -d '{
+  "name": "llama3"
+}'
+```
+
+#### Response
+
+If `stream` is not specified, or set to `true`, a stream of JSON objects is returned:
+
+The first object is the manifest:
+
+```json
+{
+  "status": "pulling manifest"
+}
+```
+
+Then there is a series of downloading responses. Until any of the download is completed, the `completed` key may not be included. The number of files to be downloaded depends on the number of layers specified in the manifest.
+
+```json
+{
+  "status": "downloading digestname",
+  "digest": "digestname",
+  "total": 2142590208,
+  "completed": 241970
+}
+```
+
+After all the files are downloaded, the final responses are:
+
+```json
+{
+    "status": "verifying sha256 digest"
+}
+{
+    "status": "writing manifest"
+}
+{
+    "status": "removing any unused layers"
+}
+{
+    "status": "success"
+}
+```
+
+if `stream` is set to false, then the response is a single JSON object:
+
+```json
+{
+  "status": "success"
+}
+```
+
+## Push a Model
+
+```shell
+POST /api/push
+```
+
+Upload a model to a model library. Requires registering for ollama.ai and adding a public key first.
+
+### Parameters
+
+- `name`: name of the model to push in the form of `<namespace>/<model>:<tag>`
+- `insecure`: (optional) allow insecure connections to the library. Only use this if you are pushing to your library during development.
+- `stream`: (optional) if `false` the response will be returned as a single response object, rather than a stream of objects
+
+### Examples
+
+#### Request
+
+```shell
+curl http://localhost:11434/api/push -d '{
+  "name": "mattw/pygmalion:latest"
+}'
+```
+
+#### Response
+
+If `stream` is not specified, or set to `true`, a stream of JSON objects is returned:
+
+```json
+{ "status": "retrieving manifest" }
+```
+
+and then:
+
+```json
+{
+  "status": "starting upload",
+  "digest": "sha256:bc07c81de745696fdf5afca05e065818a8149fb0c77266fb584d9b2cba3711ab",
+  "total": 1928429856
+}
+```
+
+Then there is a series of uploading responses:
+
+```json
+{
+  "status": "starting upload",
+  "digest": "sha256:bc07c81de745696fdf5afca05e065818a8149fb0c77266fb584d9b2cba3711ab",
+  "total": 1928429856
+}
+```
+
+Finally, when the upload is complete:
+
+```json
+{"status":"pushing manifest"}
+{"status":"success"}
+```
+
+If `stream` is set to `false`, then the response is a single JSON object:
+
+```json
+{ "status": "success" }
+```
+
+## Generate Embeddings
+
+```shell
+POST /api/embeddings
+```
+
+Generate embeddings from a model
+
+### Parameters
+
+- `model`: name of model to generate embeddings from
+- `prompt`: text to generate embeddings for
+
+Advanced parameters:
+
+- `options`: additional model parameters listed in the documentation for the [Modelfile](./modelfile.md#valid-parameters-and-values) such as `temperature`
+- `keep_alive`: controls how long the model will stay loaded into memory following the request (default: `5m`)
+
+### Examples
+
+#### Request
+
+```shell
+curl http://localhost:11434/api/embeddings -d '{
+  "model": "all-minilm",
+  "prompt": "Here is an article about llamas..."
+}'
+```
+
+#### Response
+
+```json
+{
+  "embedding": [
+    0.5670403838157654, 0.009260174818336964, 0.23178744316101074, -0.2916173040866852, -0.8924556970596313,
+    0.8785552978515625, -0.34576427936553955, 0.5742510557174683, -0.04222835972905159, -0.137906014919281
+  ]
+}
+```
+
+## List Running Models
+```shell
+GET /api/ps
+```
+
+List models that are currently loaded into memory.
+
+#### Examples
+
+### Request
+
+```shell
+curl http://localhost:11434/api/ps
+```
+
+#### Response
+
+A single JSON object will be returned.
+
+```json
+{
+  "models": [
+    {
+      "name": "mistral:latest",
+      "model": "mistral:latest",
+      "size": 5137025024,
+      "digest": "2ae6f6dd7a3dd734790bbbf58b8909a606e0e7e97e94b7604e0aa7ae4490e6d8",
+      "details": {
+        "parent_model": "",
+        "format": "gguf",
+        "family": "llama",
+        "families": [
+          "llama"
+        ],
+        "parameter_size": "7.2B",
+        "quantization_level": "Q4_0"
+      },
+      "expires_at": "2024-06-04T14:38:31.83753-07:00",
+      "size_vram": 5137025024
+    }
+  ]
+}
+```

ollama/docs/development.md

+# Development
+
+Install required tools:
+
+- cmake version 3.24 or higher
+- go version 1.22 or higher
+- gcc version 11.4.0 or higher
+
+### MacOS
+
+```bash
+brew install go cmake gcc
+```
+
+Optionally enable debugging and more verbose logging:
+
+```bash
+# At build time
+export CGO_CFLAGS="-g"
+
+# At runtime
+export OLLAMA_DEBUG=1
+```
+
+Get the required libraries and build the native LLM code:
+
+```bash
+go generate ./...
+```
+
+Then build ollama:
+
+```bash
+go build .
+```
+
+Now you can run `ollama`:
+
+```bash
+./ollama
+```
+
+### Linux
+
+#### Linux CUDA (NVIDIA)
+
+_Your operating system distribution may already have packages for NVIDIA CUDA. Distro packages are often preferable, but instructions are distro-specific. Please consult distro-specific docs for dependencies if available!_
+
+Install `cmake` and `golang` as well as [NVIDIA CUDA](https://developer.nvidia.com/cuda-downloads)
+development and runtime packages.
+
+Typically the build scripts will auto-detect CUDA, however, if your Linux distro
+or installation approach uses unusual paths, you can specify the location by
+specifying an environment variable `CUDA_LIB_DIR` to the location of the shared
+libraries, and `CUDACXX` to the location of the nvcc compiler. You can customize
+a set of target CUDA architectures by setting `CMAKE_CUDA_ARCHITECTURES` (e.g. "50;60;70")
+
+Then generate dependencies:
+
+```
+go generate ./...
+```
+
+Then build the binary:
+
+```
+go build .
+```
+
+#### Linux ROCm (AMD)
+
+_Your operating system distribution may already have packages for AMD ROCm and CLBlast. Distro packages are often preferable, but instructions are distro-specific. Please consult distro-specific docs for dependencies if available!_
+
+Install [CLBlast](https://github.com/CNugteren/CLBlast/blob/master/doc/installation.md) and [ROCm](https://rocm.docs.amd.com/en/latest/) development packages first, as well as `cmake` and `golang`.
+
+Typically the build scripts will auto-detect ROCm, however, if your Linux distro
+or installation approach uses unusual paths, you can specify the location by
+specifying an environment variable `ROCM_PATH` to the location of the ROCm
+install (typically `/opt/rocm`), and `CLBlast_DIR` to the location of the
+CLBlast install (typically `/usr/lib/cmake/CLBlast`). You can also customize
+the AMD GPU targets by setting AMDGPU_TARGETS (e.g. `AMDGPU_TARGETS="gfx1101;gfx1102"`)
+
+```
+go generate ./...
+```
+
+Then build the binary:
+
+```
+go build .
+```
+
+ROCm requires elevated privileges to access the GPU at runtime. On most distros you can add your user account to the `render` group, or run as root.
+
+#### Advanced CPU Settings
+
+By default, running `go generate ./...` will compile a few different variations
+of the LLM library based on common CPU families and vector math capabilities,
+including a lowest-common-denominator which should run on almost any 64 bit CPU
+somewhat slowly. At runtime, Ollama will auto-detect the optimal variation to
+load. If you would like to build a CPU-based build customized for your
+processor, you can set `OLLAMA_CUSTOM_CPU_DEFS` to the llama.cpp flags you would
+like to use. For example, to compile an optimized binary for an Intel i9-9880H,
+you might use:
+
+```
+OLLAMA_CUSTOM_CPU_DEFS="-DLLAMA_AVX=on -DLLAMA_AVX2=on -DLLAMA_F16C=on -DLLAMA_FMA=on" go generate ./...
+go build .
+```
+
+#### Containerized Linux Build
+
+If you have Docker available, you can build linux binaries with `./scripts/build_linux.sh` which has the CUDA and ROCm dependencies included. The resulting binary is placed in `./dist`
+
+### Windows
+
+Note: The Windows build for Ollama is still under development.
+
+First, install required tools:
+
+- MSVC toolchain - C/C++ and cmake as minimal requirements
+- Go version 1.22 or higher
+- MinGW (pick one variant) with GCC.
+  - [MinGW-w64](https://www.mingw-w64.org/)
+  - [MSYS2](https://www.msys2.org/)
+- The `ThreadJob` Powershell module: `Install-Module -Name ThreadJob -Scope CurrentUser`
+
+Then, build the `ollama` binary:
+
+```powershell
+$env:CGO_ENABLED="1"
+go generate ./...
+go build .
+```
+
+#### Windows CUDA (NVIDIA)
+
+In addition to the common Windows development tools described above, install CUDA after installing MSVC.
+
+- [NVIDIA CUDA](https://docs.nvidia.com/cuda/cuda-installation-guide-microsoft-windows/index.html)
+
+
+#### Windows ROCm (AMD Radeon)
+
+In addition to the common Windows development tools described above, install AMDs HIP package after installing MSVC.
+
+- [AMD HIP](https://www.amd.com/en/developer/resources/rocm-hub/hip-sdk.html)
+- [Strawberry Perl](https://strawberryperl.com/)
+
+Lastly, add `ninja.exe` included with MSVC to the system path (e.g. `C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\Common7\IDE\CommonExtensions\Microsoft\CMake\Ninja`).

ollama/docs/docker.md

+# Ollama Docker image
+
+### CPU only
+
+```bash
+docker run -d -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama
+```
+
+### Nvidia GPU
+Install the [NVIDIA Container Toolkit](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html#installation).
+
+#### Install with Apt
+1.  Configure the repository
+```bash
+curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey \
+    | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg
+curl -s -L https://nvidia.github.io/libnvidia-container/stable/deb/nvidia-container-toolkit.list \
+    | sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' \
+    | sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list
+sudo apt-get update
+```
+2.  Install the NVIDIA Container Toolkit packages
+```bash
+sudo apt-get install -y nvidia-container-toolkit
+```
+
+#### Install with Yum or Dnf
+1.  Configure the repository
+    
+```bash
+curl -s -L https://nvidia.github.io/libnvidia-container/stable/rpm/nvidia-container-toolkit.repo \
+    | sudo tee /etc/yum.repos.d/nvidia-container-toolkit.repo
+```
+    
+2. Install the NVIDIA Container Toolkit packages
+    
+```bash
+sudo yum install -y nvidia-container-toolkit
+```
+
+#### Configure Docker to use Nvidia driver 
+```
+sudo nvidia-ctk runtime configure --runtime=docker
+sudo systemctl restart docker
+```
+
+#### Start the container
+
+```bash
+docker run -d --gpus=all -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama
+```
+
+### AMD GPU
+
+To run Ollama using Docker with AMD GPUs, use the `rocm` tag and the following command:
+
+```
+docker run -d --device /dev/kfd --device /dev/dri -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama:rocm
+```
+
+### Run model locally
+
+Now you can run a model:
+
+```
+docker exec -it ollama ollama run llama3
+```
+
+### Try different models
+
+More models can be found on the [Ollama library](https://ollama.com/library).

ollama/docs/faq.md

+# FAQ
+
+## How can I upgrade Ollama?
+
+Ollama on macOS and Windows will automatically download updates. Click on the taskbar or menubar item and then click "Restart to update" to apply the update. Updates can also be installed by downloading the latest version [manually](https://ollama.com/download/).
+
+On Linux, re-run the install script:
+
+```shell
+curl -fsSL https://ollama.com/install.sh | sh
+```
+
+## How can I view the logs?
+
+Review the [Troubleshooting](./troubleshooting.md) docs for more about using logs.
+
+## Is my GPU compatible with Ollama?
+
+Please refer to the [GPU docs](./gpu.md).
+
+## How can I specify the context window size?
+
+By default, Ollama uses a context window size of 2048 tokens.
+
+To change this when using `ollama run`, use `/set parameter`:
+
+```
+/set parameter num_ctx 4096
+```
+
+When using the API, specify the `num_ctx` parameter:
+
+```shell
+curl http://localhost:11434/api/generate -d '{
+  "model": "llama3",
+  "prompt": "Why is the sky blue?",
+  "options": {
+    "num_ctx": 4096
+  }
+}'
+```
+
+## How can I tell if my model was loaded onto the GPU?
+
+Use the `ollama ps` command to see what models are currently loaded into memory.
+
+```shell
+ollama ps
+NAME      	ID          	SIZE 	PROCESSOR	UNTIL
+llama3:70b	bcfb190ca3a7	42 GB	100% GPU 	4 minutes from now
+```
+
+The `Processor` column will show which memory the model was loaded in to:
+* `100% GPU` means the model was loaded entirely into the GPU
+* `100% CPU` means the model was loaded entirely in system memory
+* `48%/52% CPU/GPU` means the model was loaded partially onto both the GPU and into system memory
+
+## How do I configure Ollama server?
+
+Ollama server can be configured with environment variables.
+
+### Setting environment variables on Mac
+
+If Ollama is run as a macOS application, environment variables should be set using `launchctl`:
+
+1. For each environment variable, call `launchctl setenv`.
+
+    ```bash
+    launchctl setenv OLLAMA_HOST "0.0.0.0"
+    ```
+
+2. Restart Ollama application.
+
+### Setting environment variables on Linux
+
+If Ollama is run as a systemd service, environment variables should be set using `systemctl`:
+
+1. Edit the systemd service by calling `systemctl edit ollama.service`. This will open an editor.
+
+2. For each environment variable, add a line `Environment` under section `[Service]`:
+
+    ```ini
+    [Service]
+    Environment="OLLAMA_HOST=0.0.0.0"
+    ```
+
+3. Save and exit.
+
+4. Reload `systemd` and restart Ollama:
+
+   ```bash
+   systemctl daemon-reload
+   systemctl restart ollama
+   ```
+
+### Setting environment variables on Windows
+
+On Windows, Ollama inherits your user and system environment variables.
+
+1. First Quit Ollama by clicking on it in the task bar.
+
+2. Start the Settings (Windows 11) or Control Panel (Windows 10) application and search for _environment variables_.
+
+3. Click on _Edit environment variables for your account_.
+
+4. Edit or create a new variable for your user account for `OLLAMA_HOST`, `OLLAMA_MODELS`, etc.
+
+5. Click OK/Apply to save.
+
+6. Start the Ollama application from the Windows Start menu.
+
+## How do I use Ollama behind a proxy?
+
+Ollama is compatible with proxy servers if `HTTP_PROXY` or `HTTPS_PROXY` are configured. When using either variables, ensure it is set where `ollama serve` can access the values. When using `HTTPS_PROXY`, ensure the proxy certificate is installed as a system certificate. Refer to the section above for how to use environment variables on your platform.
+
+### How do I use Ollama behind a proxy in Docker?
+
+The Ollama Docker container image can be configured to use a proxy by passing `-e HTTPS_PROXY=https://proxy.example.com` when starting the container.
+
+Alternatively, the Docker daemon can be configured to use a proxy. Instructions are available for Docker Desktop on [macOS](https://docs.docker.com/desktop/settings/mac/#proxies), [Windows](https://docs.docker.com/desktop/settings/windows/#proxies), and [Linux](https://docs.docker.com/desktop/settings/linux/#proxies), and Docker [daemon with systemd](https://docs.docker.com/config/daemon/systemd/#httphttps-proxy).
+
+Ensure the certificate is installed as a system certificate when using HTTPS. This may require a new Docker image when using a self-signed certificate.
+
+```dockerfile
+FROM ollama/ollama
+COPY my-ca.pem /usr/local/share/ca-certificates/my-ca.crt
+RUN update-ca-certificates
+```
+
+Build and run this image:
+
+```shell
+docker build -t ollama-with-ca .
+docker run -d -e HTTPS_PROXY=https://my.proxy.example.com -p 11434:11434 ollama-with-ca
+```
+
+## Does Ollama send my prompts and answers back to ollama.com?
+
+No. Ollama runs locally, and conversation data does not leave your machine.
+
+## How can I expose Ollama on my network?
+
+Ollama binds 127.0.0.1 port 11434 by default. Change the bind address with the `OLLAMA_HOST` environment variable.
+
+Refer to the section [above](#how-do-i-configure-ollama-server) for how to set environment variables on your platform.
+
+## How can I use Ollama with a proxy server?
+
+Ollama runs an HTTP server and can be exposed using a proxy server such as Nginx. To do so, configure the proxy to forward requests and optionally set required headers (if not exposing Ollama on the network). For example, with Nginx:
+
+```
+server {
+    listen 80;
+    server_name example.com;  # Replace with your domain or IP
+    location / {
+        proxy_pass http://localhost:11434;
+        proxy_set_header Host localhost:11434;
+    }
+}
+```
+
+## How can I use Ollama with ngrok?
+
+Ollama can be accessed using a range of tools for tunneling tools. For example with Ngrok:
+
+```shell
+ngrok http 11434 --host-header="localhost:11434"
+```
+
+## How can I use Ollama with Cloudflare Tunnel?
+
+To use Ollama with Cloudflare Tunnel, use the `--url` and `--http-host-header` flags:
+
+```shell
+cloudflared tunnel --url http://localhost:11434 --http-host-header="localhost:11434"
+```
+
+## How can I allow additional web origins to access Ollama?
+
+Ollama allows cross-origin requests from `127.0.0.1` and `0.0.0.0` by default. Additional origins can be configured with `OLLAMA_ORIGINS`.
+
+Refer to the section [above](#how-do-i-configure-ollama-server) for how to set environment variables on your platform.
+
+## Where are models stored?
+
+- macOS: `~/.ollama/models`
+- Linux: `/usr/share/ollama/.ollama/models`
+- Windows: `C:\Users\%username%\.ollama\models`
+
+### How do I set them to a different location?
+
+If a different directory needs to be used, set the environment variable `OLLAMA_MODELS` to the chosen directory.
+
+Refer to the section [above](#how-do-i-configure-ollama-server) for how to set environment variables on your platform.
+
+## How can I use Ollama in Visual Studio Code?
+
+There is already a large collection of plugins available for VSCode as well as other editors that leverage Ollama. See the list of [extensions & plugins](https://github.com/ollama/ollama#extensions--plugins) at the bottom of the main repository readme.
+
+## How do I use Ollama with GPU acceleration in Docker?
+
+The Ollama Docker container can be configured with GPU acceleration in Linux or Windows (with WSL2). This requires the [nvidia-container-toolkit](https://github.com/NVIDIA/nvidia-container-toolkit). See [ollama/ollama](https://hub.docker.com/r/ollama/ollama) for more details.
+
+GPU acceleration is not available for Docker Desktop in macOS due to the lack of GPU passthrough and emulation.
+
+## Why is networking slow in WSL2 on Windows 10?
+
+This can impact both installing Ollama, as well as downloading models.
+
+Open `Control Panel > Networking and Internet > View network status and tasks` and click on `Change adapter settings` on the left panel. Find the `vEthernel (WSL)` adapter, right click and select `Properties`.
+Click on `Configure` and open the `Advanced` tab. Search through each of the properties until you find `Large Send Offload Version 2 (IPv4)` and `Large Send Offload Version 2 (IPv6)`. *Disable* both of these
+properties.
+
+## How can I preload a model into Ollama to get faster response times?
+
+If you are using the API you can preload a model by sending the Ollama server an empty request. This works with both the `/api/generate` and `/api/chat` API endpoints.
+
+To preload the mistral model using the generate endpoint, use:
+```shell
+curl http://localhost:11434/api/generate -d '{"model": "mistral"}'
+```
+
+To use the chat completions endpoint, use:
+```shell
+curl http://localhost:11434/api/chat -d '{"model": "mistral"}'
+```
+
+To preload a model using the CLI, use the command:
+```shell
+ollama run llama3 ""
+```
+
+## How do I keep a model loaded in memory or make it unload immediately?
+
+By default models are kept in memory for 5 minutes before being unloaded. This allows for quicker response times if you are making numerous requests to the LLM. You may, however, want to free up the memory before the 5 minutes have elapsed or keep the model loaded indefinitely. Use the `keep_alive` parameter with either the `/api/generate` and `/api/chat` API endpoints to control how long the model is left in memory.
+
+The `keep_alive` parameter can be set to:
+* a duration string (such as "10m" or "24h")
+* a number in seconds (such as 3600)
+* any negative number which will keep the model loaded in memory (e.g. -1 or "-1m")
+* '0' which will unload the model immediately after generating a response
+
+For example, to preload a model and leave it in memory use:
+```shell
+curl http://localhost:11434/api/generate -d '{"model": "llama3", "keep_alive": -1}'
+```
+
+To unload the model and free up memory use:
+```shell
+curl http://localhost:11434/api/generate -d '{"model": "llama3", "keep_alive": 0}'
+```
+
+Alternatively, you can change the amount of time all models are loaded into memory by setting the `OLLAMA_KEEP_ALIVE` environment variable when starting the Ollama server. The `OLLAMA_KEEP_ALIVE` variable uses the same parameter types as the `keep_alive` parameter types mentioned above. Refer to section explaining [how to configure the Ollama server](#how-do-i-configure-ollama-server) to correctly set the environment variable.
+
+If you wish to override the `OLLAMA_KEEP_ALIVE` setting, use the `keep_alive` API parameter with the `/api/generate` or `/api/chat` API endpoints.
+
+## How do I manage the maximum number of requests the Ollama server can queue?
+
+If too many requests are sent to the server, it will respond with a 503 error indicating the server is overloaded.  You can adjust how many requests may be queue by setting `OLLAMA_MAX_QUEUE`.

ollama/docs/gpu.md

+# GPU
+## Nvidia
+Ollama supports Nvidia GPUs with compute capability 5.0+.
+
+Check your compute compatibility to see if your card is supported:
+[https://developer.nvidia.com/cuda-gpus](https://developer.nvidia.com/cuda-gpus)
+
+| Compute Capability | Family              | Cards                                                                                                       |
+| ------------------ | ------------------- | ----------------------------------------------------------------------------------------------------------- |
+| 9.0                | NVIDIA              | `H100`                                                                                                      |
+| 8.9                | GeForce RTX 40xx    | `RTX 4090` `RTX 4080 SUPER` `RTX 4080` `RTX 4070 Ti SUPER` `RTX 4070 Ti` `RTX 4070 SUPER` `RTX 4070` `RTX 4060 Ti` `RTX 4060`  |
+|                    | NVIDIA Professional | `L4` `L40` `RTX 6000`                                                                                       |
+| 8.6                | GeForce RTX 30xx    | `RTX 3090 Ti` `RTX 3090` `RTX 3080 Ti` `RTX 3080` `RTX 3070 Ti` `RTX 3070` `RTX 3060 Ti` `RTX 3060`         |
+|                    | NVIDIA Professional | `A40` `RTX A6000` `RTX A5000` `RTX A4000` `RTX A3000` `RTX A2000` `A10` `A16` `A2`                          |
+| 8.0                | NVIDIA              | `A100` `A30`                                                                                                |
+| 7.5                | GeForce GTX/RTX     | `GTX 1650 Ti` `TITAN RTX` `RTX 2080 Ti` `RTX 2080` `RTX 2070` `RTX 2060`                                    |
+|                    | NVIDIA Professional | `T4` `RTX 5000` `RTX 4000` `RTX 3000` `T2000` `T1200` `T1000` `T600` `T500`                                 |
+|                    | Quadro              | `RTX 8000` `RTX 6000` `RTX 5000` `RTX 4000`                                                                 |
+| 7.0                | NVIDIA              | `TITAN V` `V100` `Quadro GV100`                                                                             |
+| 6.1                | NVIDIA TITAN        | `TITAN Xp` `TITAN X`                                                                                        |
+|                    | GeForce GTX         | `GTX 1080 Ti` `GTX 1080` `GTX 1070 Ti` `GTX 1070` `GTX 1060` `GTX 1050`                                     |
+|                    | Quadro              | `P6000` `P5200` `P4200` `P3200` `P5000` `P4000` `P3000` `P2200` `P2000` `P1000` `P620` `P600` `P500` `P520` |
+|                    | Tesla               | `P40` `P4`                                                                                                  |
+| 6.0                | NVIDIA              | `Tesla P100` `Quadro GP100`                                                                                 |
+| 5.2                | GeForce GTX         | `GTX TITAN X` `GTX 980 Ti` `GTX 980` `GTX 970` `GTX 960` `GTX 950`                                          |
+|                    | Quadro              | `M6000 24GB` `M6000` `M5000` `M5500M` `M4000` `M2200` `M2000` `M620`                                        |
+|                    | Tesla               | `M60` `M40`                                                                                                 |
+| 5.0                | GeForce GTX         | `GTX 750 Ti` `GTX 750` `NVS 810`                                                                            |
+|                    | Quadro              | `K2200` `K1200` `K620` `M1200` `M520` `M5000M` `M4000M` `M3000M` `M2000M` `M1000M` `K620M` `M600M` `M500M`  |
+
+
+### GPU Selection
+
+If you have multiple NVIDIA GPUs in your system and want to limit Ollama to use
+a subset, you can set `CUDA_VISIBLE_DEVICES` to a comma separated list of GPUs.
+Numeric IDs may be used, however ordering may vary, so UUIDs are more reliable.
+You can discover the UUID of your GPUs by running `nvidia-smi -L` If you want to
+ignore the GPUs and force CPU usage, use an invalid GPU ID (e.g., "-1")
+
+### Laptop Suspend Resume
+
+On linux, after a suspend/resume cycle, sometimes Ollama will fail to discover
+your NVIDIA GPU, and fallback to running on the CPU.  You can workaround this
+driver bug by reloading the NVIDIA UVM driver with `sudo rmmod nvidia_uvm &&
+sudo modprobe nvidia_uvm`
+
+## AMD Radeon
+Ollama supports the following AMD GPUs:
+| Family         | Cards and accelerators                                                                                                               |
+| -------------- | ---------------------------------------------------------------------------------------------------------------------------------------------- |
+| AMD Radeon RX  | `7900 XTX` `7900 XT` `7900 GRE` `7800 XT` `7700 XT` `7600 XT` `7600` `6950 XT` `6900 XTX` `6900XT` `6800 XT` `6800` `Vega 64` `Vega 56`    |
+| AMD Radeon PRO | `W7900` `W7800` `W7700` `W7600` `W7500` `W6900X` `W6800X Duo` `W6800X` `W6800` `V620` `V420` `V340` `V320` `Vega II Duo` `Vega II` `VII` `SSG` |
+| AMD Instinct   | `MI300X` `MI300A` `MI300` `MI250X` `MI250` `MI210` `MI200` `MI100` `MI60` `MI50`                                                               |
+
+### Overrides
+Ollama leverages the AMD ROCm library, which does not support all AMD GPUs. In
+some cases you can force the system to try to use a similar LLVM target that is
+close.  For example The Radeon RX 5400 is `gfx1034` (also known as 10.3.4)
+however, ROCm does not currently support this target. The closest support is
+`gfx1030`.  You can use the environment variable `HSA_OVERRIDE_GFX_VERSION` with
+`x.y.z` syntax.  So for example, to force the system to run on the RX 5400, you
+would set `HSA_OVERRIDE_GFX_VERSION="10.3.0"` as an environment variable for the
+server.  If you have an unsupported AMD GPU you can experiment using the list of
+supported types below.
+
+At this time, the known supported GPU types are the following LLVM Targets.
+This table shows some example GPUs that map to these LLVM targets:
+| **LLVM Target** | **An Example GPU** |
+|-----------------|---------------------|
+| gfx900 | Radeon RX Vega 56 |
+| gfx906 | Radeon Instinct MI50 |
+| gfx908 | Radeon Instinct MI100 |
+| gfx90a | Radeon Instinct MI210 |
+| gfx940 | Radeon Instinct MI300 |
+| gfx941 | |
+| gfx942 | |
+| gfx1030 | Radeon PRO V620 |
+| gfx1100 | Radeon PRO W7900 |
+| gfx1101 | Radeon PRO W7700 |
+| gfx1102 | Radeon RX 7600 |
+
+AMD is working on enhancing ROCm v6 to broaden support for families of GPUs in a
+future release which should increase support for more GPUs.
+
+Reach out on [Discord](https://discord.gg/ollama) or file an
+[issue](https://github.com/ollama/ollama/issues) for additional help.
+
+### GPU Selection
+
+If you have multiple AMD GPUs in your system and want to limit Ollama to use a
+subset, you can set `HIP_VISIBLE_DEVICES` to a comma separated list of GPUs.
+You can see the list of devices with `rocminfo`.  If you want to ignore the GPUs
+and force CPU usage, use an invalid GPU ID (e.g., "-1")
+
+### Container Permission
+
+In some Linux distributions, SELinux can prevent containers from
+accessing the AMD GPU devices.  On the host system you can run 
+`sudo setsebool container_use_devices=1` to allow containers to use devices.
+
+### Metal (Apple GPUs)
+Ollama supports GPU acceleration on Apple devices via the Metal API.

ollama/docs/import.md

+# Import
+
+GGUF models and select Safetensors models can be imported directly into Ollama.
+
+## Import GGUF
+
+A binary GGUF file can be imported directly into Ollama through a Modelfile.
+
+```dockerfile
+FROM /path/to/file.gguf
+```
+
+## Import Safetensors
+
+If the model being imported is one of these architectures, it can be imported directly into Ollama through a Modelfile:
+
+ - LlamaForCausalLM
+ - MistralForCausalLM
+ - GemmaForCausalLM
+
+```dockerfile
+FROM /path/to/safetensors/directory
+```
+
+For architectures not directly convertable by Ollama, see llama.cpp's [guide](https://github.com/ggerganov/llama.cpp/blob/master/README.md#prepare-and-quantize) on conversion. After conversion, see [Import GGUF](#import-gguf).
+
+## Automatic Quantization
+
+> [!NOTE]
+> Automatic quantization requires v0.1.35 or higher.
+
+Ollama is capable of quantizing FP16 or FP32 models to any of the supported quantizations with the `-q/--quantize` flag in `ollama create`.
+
+```dockerfile
+FROM /path/to/my/gemma/f16/model
+```
+
+```shell
+$ ollama create -q Q4_K_M mymodel
+transferring model data
+quantizing F16 model to Q4_K_M
+creating new layer sha256:735e246cc1abfd06e9cdcf95504d6789a6cd1ad7577108a70d9902fef503c1bd
+creating new layer sha256:0853f0ad24e5865173bbf9ffcc7b0f5d56b66fd690ab1009867e45e7d2c4db0f
+writing manifest
+success
+```
+
+### Supported Quantizations
+
+- `Q4_0`
+- `Q4_1`
+- `Q5_0`
+- `Q5_1`
+- `Q8_0`
+
+#### K-means Quantizations
+
+- `Q3_K_S`
+- `Q3_K_M`
+- `Q3_K_L`
+- `Q4_K_S`
+- `Q4_K_M`
+- `Q5_K_S`
+- `Q5_K_M`
+- `Q6_K`
+
+## Template Detection
+
+> [!NOTE]
+> Template detection requires v0.1.42 or higher.
+
+Ollama uses model metadata, specifically `tokenizer.chat_template`, to automatically create a template appropriate for the model you're importing.
+
+```dockerfile
+FROM /path/to/my/gemma/model
+```
+
+```shell
+$ ollama create mymodel
+transferring model data
+using autodetected template gemma-instruct
+creating new layer sha256:baa2a0edc27d19cc6b7537578a9a7ba1a4e3214dc185ed5ae43692b319af7b84
+creating new layer sha256:ba66c3309914dbef07e5149a648fd1877f030d337a4f240d444ea335008943cb
+writing manifest
+success
+```
+
+Defining a template in the Modelfile will disable this feature which may be useful if you want to use a different template than the autodetected one.

ollama/docs/linux.md

+# Ollama on Linux
+
+## Install
+
+Install Ollama running this one-liner:
+
+>
+
+```bash
+curl -fsSL https://ollama.com/install.sh | sh
+```
+
+## AMD Radeon GPU support
+
+While AMD has contributed the `amdgpu` driver upstream to the official linux
+kernel source, the version is older and may not support all ROCm features. We
+recommend you install the latest driver from
+https://www.amd.com/en/support/linux-drivers for best support of your Radeon
+GPU.
+
+## Manual install
+
+### Download the `ollama` binary
+
+Ollama is distributed as a self-contained binary. Download it to a directory in your PATH:
+
+```bash
+sudo curl -L https://ollama.com/download/ollama-linux-amd64 -o /usr/bin/ollama
+sudo chmod +x /usr/bin/ollama
+```
+
+### Adding Ollama as a startup service (recommended)
+
+Create a user for Ollama:
+
+```bash
+sudo useradd -r -s /bin/false -m -d /usr/share/ollama ollama
+```
+
+Create a service file in `/etc/systemd/system/ollama.service`:
+
+```ini
+[Unit]
+Description=Ollama Service
+After=network-online.target
+
+[Service]
+ExecStart=/usr/bin/ollama serve
+User=ollama
+Group=ollama
+Restart=always
+RestartSec=3
+
+[Install]
+WantedBy=default.target
+```
+
+Then start the service:
+
+```bash
+sudo systemctl daemon-reload
+sudo systemctl enable ollama
+```
+
+### Install CUDA drivers (optional – for Nvidia GPUs)
+
+[Download and install](https://developer.nvidia.com/cuda-downloads) CUDA.
+
+Verify that the drivers are installed by running the following command, which should print details about your GPU:
+
+```bash
+nvidia-smi
+```
+
+### Install ROCm (optional - for Radeon GPUs)
+[Download and Install](https://rocm.docs.amd.com/projects/install-on-linux/en/latest/tutorial/quick-start.html)
+
+Make sure to install ROCm v6
+
+### Start Ollama
+
+Start Ollama using `systemd`:
+
+```bash
+sudo systemctl start ollama
+```
+
+## Update
+
+Update ollama by running the install script again:
+
+```bash
+curl -fsSL https://ollama.com/install.sh | sh
+```
+
+Or by downloading the ollama binary:
+
+```bash
+sudo curl -L https://ollama.com/download/ollama-linux-amd64 -o /usr/bin/ollama
+sudo chmod +x /usr/bin/ollama
+```
+
+## Installing specific versions
+
+Use `OLLAMA_VERSION` environment variable with the install script to install a specific version of Ollama, including pre-releases. You can find the version numbers in the [releases page](https://github.com/ollama/ollama/releases). 
+
+For example:
+
+```
+curl -fsSL https://ollama.com/install.sh | OLLAMA_VERSION=0.1.32 sh
+```
+
+## Viewing logs
+
+To view logs of Ollama running as a startup service, run:
+
+```bash
+journalctl -e -u ollama
+```
+
+## Uninstall
+
+Remove the ollama service:
+
+```bash
+sudo systemctl stop ollama
+sudo systemctl disable ollama
+sudo rm /etc/systemd/system/ollama.service
+```
+
+Remove the ollama binary from your bin directory (either `/usr/local/bin`, `/usr/bin`, or `/bin`):
+
+```bash
+sudo rm $(which ollama)
+```
+
+Remove the downloaded models and Ollama service user and group:
+
+```bash
+sudo rm -r /usr/share/ollama
+sudo userdel ollama
+sudo groupdel ollama
+```

ollama/docs/modelfile.md

+# Ollama Model File
+
+> Note: `Modelfile` syntax is in development
+
+A model file is the blueprint to create and share models with Ollama.
+
+## Table of Contents
+
+- [Format](#format)
+- [Examples](#examples)
+- [Instructions](#instructions)
+  - [FROM (Required)](#from-required)
+    - [Build from llama3](#build-from-llama3)
+    - [Build from a bin file](#build-from-a-bin-file)
+  - [PARAMETER](#parameter)
+    - [Valid Parameters and Values](#valid-parameters-and-values)
+  - [TEMPLATE](#template)
+    - [Template Variables](#template-variables)
+  - [SYSTEM](#system)
+  - [ADAPTER](#adapter)
+  - [LICENSE](#license)
+  - [MESSAGE](#message)
+- [Notes](#notes)
+
+## Format
+
+The format of the `Modelfile`:
+
+```modelfile
+# comment
+INSTRUCTION arguments
+```
+
+| Instruction                         | Description                                                    |
+| ----------------------------------- | -------------------------------------------------------------- |
+| [`FROM`](#from-required) (required) | Defines the base model to use.                                 |
+| [`PARAMETER`](#parameter)           | Sets the parameters for how Ollama will run the model.         |
+| [`TEMPLATE`](#template)             | The full prompt template to be sent to the model.              |
+| [`SYSTEM`](#system)                 | Specifies the system message that will be set in the template. |
+| [`ADAPTER`](#adapter)               | Defines the (Q)LoRA adapters to apply to the model.            |
+| [`LICENSE`](#license)               | Specifies the legal license.                                   |
+| [`MESSAGE`](#message)               | Specify message history.                                       |
+
+## Examples
+
+### Basic `Modelfile`
+
+An example of a `Modelfile` creating a mario blueprint:
+
+```modelfile
+FROM llama3
+# sets the temperature to 1 [higher is more creative, lower is more coherent]
+PARAMETER temperature 1
+# sets the context window size to 4096, this controls how many tokens the LLM can use as context to generate the next token
+PARAMETER num_ctx 4096
+
+# sets a custom system message to specify the behavior of the chat assistant
+SYSTEM You are Mario from super mario bros, acting as an assistant.
+```
+
+To use this:
+
+1. Save it as a file (e.g. `Modelfile`)
+2. `ollama create choose-a-model-name -f <location of the file e.g. ./Modelfile>'`
+3. `ollama run choose-a-model-name`
+4. Start using the model!
+
+More examples are available in the [examples directory](../examples).
+
+To view the Modelfile of a given model, use the `ollama show --modelfile` command.
+
+  ```bash
+  > ollama show --modelfile llama3
+  # Modelfile generated by "ollama show"
+  # To build a new Modelfile based on this one, replace the FROM line with:
+  # FROM llama3:latest
+  FROM /Users/pdevine/.ollama/models/blobs/sha256-00e1317cbf74d901080d7100f57580ba8dd8de57203072dc6f668324ba545f29
+  TEMPLATE """{{ if .System }}<|start_header_id|>system<|end_header_id|>
+
+  {{ .System }}<|eot_id|>{{ end }}{{ if .Prompt }}<|start_header_id|>user<|end_header_id|>
+
+  {{ .Prompt }}<|eot_id|>{{ end }}<|start_header_id|>assistant<|end_header_id|>
+
+  {{ .Response }}<|eot_id|>"""
+  PARAMETER stop "<|start_header_id|>"
+  PARAMETER stop "<|end_header_id|>"
+  PARAMETER stop "<|eot_id|>"
+  PARAMETER stop "<|reserved_special_token"
+  ```
+
+## Instructions
+
+### FROM (Required)
+
+The `FROM` instruction defines the base model to use when creating a model.
+
+```modelfile
+FROM <model name>:<tag>
+```
+
+#### Build from llama3
+
+```modelfile
+FROM llama3
+```
+
+A list of available base models:
+<https://github.com/ollama/ollama#model-library>
+
+#### Build from a `bin` file
+
+```modelfile
+FROM ./ollama-model.bin
+```
+
+This bin file location should be specified as an absolute path or relative to the `Modelfile` location.
+
+### PARAMETER
+
+The `PARAMETER` instruction defines a parameter that can be set when the model is run.
+
+```modelfile
+PARAMETER <parameter> <parametervalue>
+```
+
+#### Valid Parameters and Values
+
+| Parameter      | Description                                                                                                                                                                                                                                             | Value Type | Example Usage        |
+| -------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ---------- | -------------------- |
+| mirostat       | Enable Mirostat sampling for controlling perplexity. (default: 0, 0 = disabled, 1 = Mirostat, 2 = Mirostat 2.0)                                                                                                                                         | int        | mirostat 0           |
+| mirostat_eta   | Influences how quickly the algorithm responds to feedback from the generated text. A lower learning rate will result in slower adjustments, while a higher learning rate will make the algorithm more responsive. (Default: 0.1)                        | float      | mirostat_eta 0.1     |
+| mirostat_tau   | Controls the balance between coherence and diversity of the output. A lower value will result in more focused and coherent text. (Default: 5.0)                                                                                                         | float      | mirostat_tau 5.0     |
+| num_ctx        | Sets the size of the context window used to generate the next token. (Default: 2048)                                                                                                                                                                    | int        | num_ctx 4096         |
+| repeat_last_n  | Sets how far back for the model to look back to prevent repetition. (Default: 64, 0 = disabled, -1 = num_ctx)                                                                                                                                           | int        | repeat_last_n 64     |
+| repeat_penalty | Sets how strongly to penalize repetitions. A higher value (e.g., 1.5) will penalize repetitions more strongly, while a lower value (e.g., 0.9) will be more lenient. (Default: 1.1)                                                                     | float      | repeat_penalty 1.1   |
+| temperature    | The temperature of the model. Increasing the temperature will make the model answer more creatively. (Default: 0.8)                                                                                                                                     | float      | temperature 0.7      |
+| seed           | Sets the random number seed to use for generation. Setting this to a specific number will make the model generate the same text for the same prompt. (Default: 0)                                                                                       | int        | seed 42              |
+| stop           | Sets the stop sequences to use. When this pattern is encountered the LLM will stop generating text and return. Multiple stop patterns may be set by specifying multiple separate `stop` parameters in a modelfile.                                      | string     | stop "AI assistant:" |
+| tfs_z          | Tail free sampling is used to reduce the impact of less probable tokens from the output. A higher value (e.g., 2.0) will reduce the impact more, while a value of 1.0 disables this setting. (default: 1)                                               | float      | tfs_z 1              |
+| num_predict    | Maximum number of tokens to predict when generating text. (Default: 128, -1 = infinite generation, -2 = fill context)                                                                                                                                   | int        | num_predict 42       |
+| top_k          | Reduces the probability of generating nonsense. A higher value (e.g. 100) will give more diverse answers, while a lower value (e.g. 10) will be more conservative. (Default: 40)                                                                        | int        | top_k 40             |
+| top_p          | Works together with top-k. A higher value (e.g., 0.95) will lead to more diverse text, while a lower value (e.g., 0.5) will generate more focused and conservative text. (Default: 0.9)                                                                 | float      | top_p 0.9            |
+
+### TEMPLATE
+
+`TEMPLATE` of the full prompt template to be passed into the model. It may include (optionally) a system message, a user's message and the response from the model. Note: syntax may be model specific. Templates use Go [template syntax](https://pkg.go.dev/text/template).
+
+#### Template Variables
+
+| Variable          | Description                                                                                   |
+| ----------------- | --------------------------------------------------------------------------------------------- |
+| `{{ .System }}`   | The system message used to specify custom behavior.                                           |
+| `{{ .Prompt }}`   | The user prompt message.                                                                      |
+| `{{ .Response }}` | The response from the model. When generating a response, text after this variable is omitted. |
+
+```
+TEMPLATE """{{ if .System }}<|im_start|>system
+{{ .System }}<|im_end|>
+{{ end }}{{ if .Prompt }}<|im_start|>user
+{{ .Prompt }}<|im_end|>
+{{ end }}<|im_start|>assistant
+"""
+```
+
+### SYSTEM
+
+The `SYSTEM` instruction specifies the system message to be used in the template, if applicable.
+
+```modelfile
+SYSTEM """<system message>"""
+```
+
+### ADAPTER
+
+The `ADAPTER` instruction is an optional instruction that specifies any LoRA adapter that should apply to the base model. The value of this instruction should be an absolute path or a path relative to the Modelfile and the file must be in a GGML file format. The adapter should be tuned from the base model otherwise the behaviour is undefined.
+
+```modelfile
+ADAPTER ./ollama-lora.bin
+```
+
+### LICENSE
+
+The `LICENSE` instruction allows you to specify the legal license under which the model used with this Modelfile is shared or distributed.
+
+```modelfile
+LICENSE """
+<license text>
+"""
+```
+
+### MESSAGE
+
+The `MESSAGE` instruction allows you to specify a message history for the model to use when responding. Use multiple iterations of the MESSAGE command to build up a conversation which will guide the model to answer in a similar way.
+
+```modelfile
+MESSAGE <role> <message>
+```
+
+#### Valid roles
+
+| Role      | Description                                                  |
+| --------- | ------------------------------------------------------------ |
+| system    | Alternate way of providing the SYSTEM message for the model. |
+| user      | An example message of what the user could have asked.        |
+| assistant | An example message of how the model should respond.          |
+
+
+#### Example conversation
+
+```modelfile
+MESSAGE user Is Toronto in Canada?
+MESSAGE assistant yes
+MESSAGE user Is Sacramento in Canada?
+MESSAGE assistant no
+MESSAGE user Is Ontario in Canada?
+MESSAGE assistant yes
+```
+
+
+## Notes
+
+- the **`Modelfile` is not case sensitive**. In the examples, uppercase instructions are used to make it easier to distinguish it from arguments.
+- Instructions can be in any order. In the examples, the `FROM` instruction is first to keep it easily readable.
+
+[1]: https://ollama.com/library

ollama/docs/openai.md

+# OpenAI compatibility
+
+> **Note:** OpenAI compatibility is experimental and is subject to major adjustments including breaking changes. For fully-featured access to the Ollama API, see the Ollama [Python library](https://github.com/ollama/ollama-python), [JavaScript library](https://github.com/ollama/ollama-js) and [REST API](https://github.com/ollama/ollama/blob/main/docs/api.md).
+
+Ollama provides experimental compatibility with parts of the [OpenAI API](https://platform.openai.com/docs/api-reference) to help connect existing applications to Ollama.
+
+## Usage
+
+### OpenAI Python library
+
+```python
+from openai import OpenAI
+
+client = OpenAI(
+    base_url='http://localhost:11434/v1/',
+
+    # required but ignored
+    api_key='ollama',
+)
+
+chat_completion = client.chat.completions.create(
+    messages=[
+        {
+            'role': 'user',
+            'content': 'Say this is a test',
+        }
+    ],
+    model='llama3',
+)
+```
+
+### OpenAI JavaScript library
+
+```javascript
+import OpenAI from 'openai'
+
+const openai = new OpenAI({
+  baseURL: 'http://localhost:11434/v1/',
+
+  // required but ignored
+  apiKey: 'ollama',
+})
+
+const chatCompletion = await openai.chat.completions.create({
+  messages: [{ role: 'user', content: 'Say this is a test' }],
+  model: 'llama3',
+})
+```
+
+### `curl`
+
+```
+curl http://localhost:11434/v1/chat/completions \
+    -H "Content-Type: application/json" \
+    -d '{
+        "model": "llama3",
+        "messages": [
+            {
+                "role": "system",
+                "content": "You are a helpful assistant."
+            },
+            {
+                "role": "user",
+                "content": "Hello!"
+            }
+        ]
+    }'
+```
+
+## Endpoints
+
+### `/v1/chat/completions`
+
+#### Supported features
+
+- [x] Chat completions
+- [x] Streaming
+- [x] JSON mode
+- [x] Reproducible outputs
+- [ ] Vision
+- [ ] Function calling
+- [ ] Logprobs
+
+#### Supported request fields
+
+- [x] `model`
+- [x] `messages`
+  - [x] Text `content`
+  - [ ] Array of `content` parts
+- [x] `frequency_penalty`
+- [x] `presence_penalty`
+- [x] `response_format`
+- [x] `seed`
+- [x] `stop`
+- [x] `stream`
+- [x] `temperature`
+- [x] `top_p`
+- [x] `max_tokens`
+- [ ] `logit_bias`
+- [ ] `tools`
+- [ ] `tool_choice`
+- [ ] `user`
+- [ ] `n`
+
+#### Notes
+
+- `finish_reason` will always be `stop`
+- `usage.prompt_tokens` will be 0 for completions where prompt evaluation is cached
+
+## Models
+
+Before using a model, pull it locally `ollama pull`:
+
+```shell
+ollama pull llama3
+```
+
+### Default model names
+
+For tooling that relies on default OpenAI model names such as `gpt-3.5-turbo`, use `ollama cp` to copy an existing model name to a temporary name:
+
+```
+ollama cp llama3 gpt-3.5-turbo
+```
+
+Afterwards, this new model name can be specified the `model` field:
+
+```shell
+curl http://localhost:11434/v1/chat/completions \
+    -H "Content-Type: application/json" \
+    -d '{
+        "model": "gpt-3.5-turbo",
+        "messages": [
+            {
+                "role": "user",
+                "content": "Hello!"
+            }
+        ]
+    }'
+```

ollama/docs/troubleshooting.md

+# How to troubleshoot issues
+
+Sometimes Ollama may not perform as expected. One of the best ways to figure out what happened is to take a look at the logs. Find the logs on **Mac** by running the command:
+
+```shell
+cat ~/.ollama/logs/server.log
+```
+
+On **Linux** systems with systemd, the logs can be found with this command:
+
+```shell
+journalctl -u ollama
+```
+
+When you run Ollama in a **container**, the logs go to stdout/stderr in the container:
+
+```shell
+docker logs <container-name>
+```
+(Use `docker ps` to find the container name)
+
+If manually running `ollama serve` in a terminal, the logs will be on that terminal.
+
+When you run Ollama on **Windows**, there are a few different locations. You can view them in the explorer window by hitting `<cmd>+R` and type in:
+- `explorer %LOCALAPPDATA%\Ollama` to view logs
+- `explorer %LOCALAPPDATA%\Programs\Ollama` to browse the binaries (The installer adds this to your user PATH)
+- `explorer %HOMEPATH%\.ollama` to browse where models and configuration is stored
+- `explorer %TEMP%` where temporary executable files are stored in one or more `ollama*` directories
+
+To enable additional debug logging to help troubleshoot problems, first **Quit the running app from the tray menu** then in a powershell terminal
+```powershell
+$env:OLLAMA_DEBUG="1"
+& "ollama app.exe"
+```
+
+Join the [Discord](https://discord.gg/ollama) for help interpreting the logs.
+
+## LLM libraries
+
+Ollama includes multiple LLM libraries compiled for different GPUs and CPU vector features. Ollama tries to pick the best one based on the capabilities of your system. If this autodetection has problems, or you run into other problems (e.g. crashes in your GPU) you can workaround this by forcing a specific LLM library. `cpu_avx2` will perform the best, followed by `cpu_avx` an the slowest but most compatible is `cpu`. Rosetta emulation under MacOS will work with the `cpu` library. 
+
+In the server log, you will see a message that looks something like this (varies from release to release):
+
+```
+Dynamic LLM libraries [rocm_v6 cpu cpu_avx cpu_avx2 cuda_v11 rocm_v5]
+```
+
+**Experimental LLM Library Override**
+
+You can set OLLAMA_LLM_LIBRARY to any of the available LLM libraries to bypass autodetection, so for example, if you have a CUDA card, but want to force the CPU LLM library with AVX2 vector support, use:
+
+```
+OLLAMA_LLM_LIBRARY="cpu_avx2" ollama serve
+```
+
+You can see what features your CPU has with the following.
+```
+cat /proc/cpuinfo| grep flags | head -1
+```
+
+## Installing older or pre-release versions on Linux
+
+If you run into problems on Linux and want to install an older version, or you'd like to try out a pre-release before it's officially released, you can tell the install script which version to install.
+
+```sh
+curl -fsSL https://ollama.com/install.sh | OLLAMA_VERSION="0.1.29" sh
+```
+
+## Linux tmp noexec 
+
+If your system is configured with the "noexec" flag where Ollama stores its temporary executable files, you can specify an alternate location by setting OLLAMA_TMPDIR to a location writable by the user ollama runs as. For example OLLAMA_TMPDIR=/usr/share/ollama/
+
+## Container fails to run on NVIDIA GPU
+
+Make sure you've set up the container runtime first as described in [docker.md](./docker.md)
+
+Sometimes the container runtime can have difficulties initializing the GPU. When you check the server logs, this can show up as various error codes, such as "3" (not initialized), "46" (device unavailable), "100" (no device), "999" (unknown), or others. The following troubleshooting techniques may help resolve the problem
+
+- Is the container runtime working?  Try `docker run --gpus all ubuntu nvidia-smi` - if this doesn't work, Ollama wont be able to see your NVIDIA GPU.
+- Is the uvm driver not loaded? `sudo nvidia-modprobe -u`
+- Try reloading the nvidia_uvm driver - `sudo rmmod nvidia_uvm` then `sudo modprobe nvidia_uvm`
+- Try rebooting
+- Make sure you're running the latest nvidia drivers
+
+If none of those resolve the problem, gather additional information and file an issue:
+- Set `CUDA_ERROR_LEVEL=50` and try again to get more diagnostic logs
+- Check dmesg for any errors `sudo dmesg | grep -i nvrm` and `sudo dmesg | grep -i nvidia`

ollama/docs/tutorials.md

+# Tutorials
+
+Here is a list of ways you can use Ollama with other tools to build interesting applications.
+
+- [Using LangChain with Ollama in JavaScript](./tutorials/langchainjs.md)
+- [Using LangChain with Ollama in Python](./tutorials/langchainpy.md)
+- [Running Ollama on NVIDIA Jetson Devices](./tutorials/nvidia-jetson.md)
+
+Also be sure to check out the [examples](../examples) directory for more ways to use Ollama.

ollama/docs/tutorials/fly-gpu.md

+# Running Ollama on Fly.io GPU Instances
+
+Ollama runs with little to no configuration on [Fly.io GPU instances](https://fly.io/docs/gpus/gpu-quickstart/). If you don't have access to GPUs yet, you'll need to [apply for access](https://fly.io/gpu/) on the waitlist. Once you're accepted, you'll get an email with instructions on how to get started.
+
+Create a new app with `fly apps create`:
+
+```bash
+fly apps create
+```
+
+Then create a `fly.toml` file in a new folder that looks like this:
+
+```toml
+app = "sparkling-violet-709"
+primary_region = "ord"
+vm.size = "a100-40gb" # see https://fly.io/docs/gpus/gpu-quickstart/ for more info
+
+[build]
+  image = "ollama/ollama"
+
+[http_service]
+  internal_port = 11434
+  force_https = false
+  auto_stop_machines = true
+  auto_start_machines = true
+  min_machines_running = 0
+  processes = ["app"]
+
+[mounts]
+  source = "models"
+  destination = "/root/.ollama"
+  initial_size = "100gb"
+```
+
+Then create a [new private IPv6 address](https://fly.io/docs/reference/private-networking/#flycast-private-load-balancing) for your app:
+
+```bash
+fly ips allocate-v6 --private
+```
+
+Then deploy your app:
+
+```bash
+fly deploy
+```
+
+And finally you can access it interactively with a new Fly.io Machine:
+
+```
+fly machine run -e OLLAMA_HOST=http://your-app-name.flycast --shell ollama/ollama
+```
+
+```bash
+$ ollama run openchat:7b-v3.5-fp16
+>>> How do I bake chocolate chip cookies?
+ To bake chocolate chip cookies, follow these steps:
+
+1. Preheat the oven to 375°F (190°C) and line a baking sheet with parchment paper or silicone baking mat.
+
+2. In a large bowl, mix together 1 cup of unsalted butter (softened), 3/4 cup granulated sugar, and 3/4
+cup packed brown sugar until light and fluffy.
+
+3. Add 2 large eggs, one at a time, to the butter mixture, beating well after each addition. Stir in 1
+teaspoon of pure vanilla extract.
+
+4. In a separate bowl, whisk together 2 cups all-purpose flour, 1/2 teaspoon baking soda, and 1/2 teaspoon
+salt. Gradually add the dry ingredients to the wet ingredients, stirring until just combined.
+
+5. Fold in 2 cups of chocolate chips (or chunks) into the dough.
+
+6. Drop rounded tablespoons of dough onto the prepared baking sheet, spacing them about 2 inches apart.
+
+7. Bake for 10-12 minutes, or until the edges are golden brown. The centers should still be slightly soft.
+
+8. Allow the cookies to cool on the baking sheet for a few minutes before transferring them to a wire rack
+to cool completely.
+
+Enjoy your homemade chocolate chip cookies!
+```
+
+When you set it up like this, it will automatically turn off when you're done using it. Then when you access it again, it will automatically turn back on. This is a great way to save money on GPU instances when you're not using them. If you want a persistent wake-on-use connection to your Ollama instance, you can set up a [connection to your Fly network using WireGuard](https://fly.io/docs/reference/private-networking/#discovering-apps-through-dns-on-a-wireguard-connection). Then you can access your Ollama instance at `http://your-app-name.flycast`.
+
+And that's it!

ollama/docs/tutorials/langchainjs.md

+# Using LangChain with Ollama using JavaScript
+
+In this tutorial, we are going to use JavaScript with LangChain and Ollama to learn about something just a touch more recent. In August 2023, there was a series of wildfires on Maui. There is no way an LLM trained before that time can know about this, since their training data would not include anything as recent as that. So we can find the [Wikipedia article about the fires](https://en.wikipedia.org/wiki/2023_Hawaii_wildfires) and ask questions about the contents.
+
+To get started, let's just use **LangChain** to ask a simple question to a model. To do this with JavaScript, we need to install **LangChain**:
+
+```bash
+npm install @langchain/community
+```
+
+Now we can start building out our JavaScript:
+
+```javascript
+import { Ollama } from "@langchain/community/llms/ollama";
+
+const ollama = new Ollama({
+  baseUrl: "http://localhost:11434",
+  model: "llama3",
+});
+
+const answer = await ollama.invoke(`why is the sky blue?`);
+
+console.log(answer);
+```
+
+That will get us the same thing as if we ran `ollama run llama3 "why is the sky blue"` in the terminal. But we want to load a document from the web to ask a question against. **Cheerio** is a great library for ingesting a webpage, and **LangChain** uses it in their **CheerioWebBaseLoader**. So let's install **Cheerio** and build that part of the app.
+
+```bash
+npm install cheerio
+```
+
+```javascript
+import { CheerioWebBaseLoader } from "langchain/document_loaders/web/cheerio";
+
+const loader = new CheerioWebBaseLoader("https://en.wikipedia.org/wiki/2023_Hawaii_wildfires");
+const data = await loader.load();
+```
+
+That will load the document. Although this page is smaller than the Odyssey, it is certainly bigger than the context size for most LLMs. So we are going to need to split into smaller pieces, and then select just the pieces relevant to our question. This is a great use for a vector datastore. In this example, we will use the **MemoryVectorStore** that is part of **LangChain**. But there is one more thing we need to get the content into the datastore. We have to run an embeddings process that converts the tokens in the text into a series of vectors. And for that, we are going to use **Tensorflow**. There is a lot of stuff going on in this one. First, install the **Tensorflow** components that we need.
+
+```javascript
+npm install @tensorflow/tfjs-core@3.6.0 @tensorflow/tfjs-converter@3.6.0 @tensorflow-models/universal-sentence-encoder@1.3.3 @tensorflow/tfjs-node@4.10.0
+```
+
+If you just install those components without the version numbers, it will install the latest versions, but there are conflicts within **Tensorflow**, so you need to install the compatible versions.
+
+```javascript
+import { RecursiveCharacterTextSplitter } from "langchain/text_splitter"
+import { MemoryVectorStore } from "langchain/vectorstores/memory";
+import "@tensorflow/tfjs-node";
+import { TensorFlowEmbeddings } from "langchain/embeddings/tensorflow";
+
+// Split the text into 500 character chunks. And overlap each chunk by 20 characters
+const textSplitter = new RecursiveCharacterTextSplitter({
+ chunkSize: 500,
+ chunkOverlap: 20
+});
+const splitDocs = await textSplitter.splitDocuments(data);
+
+// Then use the TensorFlow Embedding to store these chunks in the datastore
+const vectorStore = await MemoryVectorStore.fromDocuments(splitDocs, new TensorFlowEmbeddings());
+```
+
+To connect the datastore to a question asked to a LLM, we need to use the concept at the heart of **LangChain**: the chain. Chains are a way to connect a number of activities together to accomplish a particular tasks. There are a number of chain types available, but for this tutorial we are using the **RetrievalQAChain**.
+
+```javascript
+import { RetrievalQAChain } from "langchain/chains";
+
+const retriever = vectorStore.asRetriever();
+const chain = RetrievalQAChain.fromLLM(ollama, retriever);
+const result = await chain.call({query: "When was Hawaii's request for a major disaster declaration approved?"});
+console.log(result.text)
+```
+
+So we created a retriever, which is a way to return the chunks that match a query from a datastore. And then connect the retriever and the model via a chain. Finally, we send a query to the chain, which results in an answer using our document as a source. The answer it returned was correct, August 10, 2023.
+
+And that is a simple introduction to what you can do with **LangChain** and **Ollama.**

ollama/docs/tutorials/langchainpy.md

+# Using LangChain with Ollama in Python
+
+Let's imagine we are studying the classics, such as **the Odyssey** by **Homer**. We might have a question about Neleus and his family. If you ask llama2 for that info, you may get something like:
+
+> I apologize, but I'm a large language model, I cannot provide information on individuals or families that do not exist in reality. Neleus is not a real person or character, and therefore does not have a family or any other personal details. My apologies for any confusion. Is there anything else I can help you with?
+
+This sounds like a typical censored response, but even llama2-uncensored gives a mediocre answer:
+
+> Neleus was a legendary king of Pylos and the father of Nestor, one of the Argonauts. His mother was Clymene, a sea nymph, while his father was Neptune, the god of the sea.
+
+So let's figure out how we can use **LangChain** with Ollama to ask our question to the actual document, the Odyssey by Homer, using Python.
+
+Let's start by asking a simple question that we can get an answer to from the **Llama2** model using **Ollama**. First, we need to install the **LangChain** package:
+
+`pip install langchain_community`
+
+Then we can create a model and ask the question:
+
+```python
+from langchain_community.llms import Ollama
+ollama = Ollama(
+    base_url='http://localhost:11434',
+    model="llama3"
+)
+print(ollama.invoke("why is the sky blue"))
+```
+
+Notice that we are defining the model and the base URL for Ollama.
+
+Now let's load a document to ask questions against. I'll load up the Odyssey by Homer, which you can find at Project Gutenberg. We will need **WebBaseLoader** which is part of **LangChain** and loads text from any webpage. On my machine, I also needed to install **bs4** to get that to work, so run `pip install bs4`.
+
+```python
+from langchain.document_loaders import WebBaseLoader
+loader = WebBaseLoader("https://www.gutenberg.org/files/1727/1727-h/1727-h.htm")
+data = loader.load()
+```
+
+This file is pretty big. Just the preface is 3000 tokens. Which means the full document won't fit into the context for the model. So we need to split it up into smaller pieces.
+
+```python
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+
+text_splitter=RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=0)
+all_splits = text_splitter.split_documents(data)
+```
+
+It's split up, but we have to find the relevant splits and then submit those to the model. We can do this by creating embeddings and storing them in a vector database. We can use Ollama directly to instantiate an embedding model. We will use ChromaDB in this example for a vector database. `pip install chromadb`
+We also need to pull embedding model: `ollama pull nomic-embed-text`
+```python
+from langchain.embeddings import OllamaEmbeddings
+from langchain.vectorstores import Chroma
+oembed = OllamaEmbeddings(base_url="http://localhost:11434", model="nomic-embed-text")
+vectorstore = Chroma.from_documents(documents=all_splits, embedding=oembed)
+```
+
+Now let's ask a question from the document. **Who was Neleus, and who is in his family?** Neleus is a character in the Odyssey, and the answer can be found in our text.
+
+```python
+question="Who is Neleus and who is in Neleus' family?"
+docs = vectorstore.similarity_search(question)
+len(docs)
+```
+
+This will output the number of matches for chunks of data similar to the search.
+
+The next thing is to send the question and the relevant parts of the docs to the model to see if we can get a good answer. But we are stitching two parts of the process together, and that is called a chain. This means we need to define a chain:
+
+```python
+from langchain.chains import RetrievalQA
+qachain=RetrievalQA.from_chain_type(ollama, retriever=vectorstore.as_retriever())
+res = qachain.invoke({"query": question})
+print(res['result'])
+```
+
+The answer received from this chain was:
+
+> Neleus is a character in Homer's "Odyssey" and is mentioned in the context of Penelope's suitors. Neleus is the father of Chloris, who is married to Neleus and bears him several children, including Nestor, Chromius, Periclymenus, and Pero. Amphinomus, the son of Nisus, is also mentioned as a suitor of Penelope and is known for his good natural disposition and agreeable conversation.
+
+It's not a perfect answer, as it implies Neleus married his daughter when actually Chloris "was the youngest daughter to Amphion son of Iasus and king of Minyan Orchomenus, and was Queen in Pylos".
+
+I updated the chunk_overlap for the text splitter to 20 and tried again and got a much better answer:
+
+> Neleus is a character in Homer's epic poem "The Odyssey." He is the husband of Chloris, who is the youngest daughter of Amphion son of Iasus and king of Minyan Orchomenus. Neleus has several children with Chloris, including Nestor, Chromius, Periclymenus, and Pero.
+
+And that is a much better answer.

ollama/docs/tutorials/nvidia-jetson.md

+# Running Ollama on NVIDIA Jetson Devices
+
+Ollama runs well on [NVIDIA Jetson Devices](https://www.nvidia.com/en-us/autonomous-machines/embedded-systems/) and should run out of the box with the standard installation instructions. 
+
+The following has been tested on [JetPack 5.1.2](https://developer.nvidia.com/embedded/jetpack), but should also work on JetPack 6.0.
+
+- Install Ollama via standard Linux command (ignore the 404 error): `curl https://ollama.com/install.sh | sh`
+- Pull the model you want to use (e.g. mistral): `ollama pull mistral`
+- Start an interactive session: `ollama run mistral`
+
+And that's it!
+
+# Running Ollama in Docker
+
+When running GPU accelerated applications in Docker, it is highly recommended to use [dusty-nv jetson-containers repo](https://github.com/dusty-nv/jetson-containers).
\ No newline at end of file

ollama/docs/windows.md

+# Ollama Windows Preview
+
+Welcome to the Ollama Windows preview.
+
+No more WSL required!
+
+Ollama now runs as a native Windows application, including NVIDIA and AMD Radeon GPU support.
+After installing Ollama Windows Preview, Ollama will run in the background and
+the `ollama` command line is available in `cmd`, `powershell` or your favorite
+terminal application. As usual the Ollama [api](./api.md) will be served on
+`http://localhost:11434`.
+
+As this is a preview release, you should expect a few bugs here and there.  If
+you run into a problem you can reach out on
+[Discord](https://discord.gg/ollama), or file an
+[issue](https://github.com/ollama/ollama/issues).
+Logs will often be helpful in diagnosing the problem (see
+[Troubleshooting](#troubleshooting) below)
+
+## System Requirements
+
+* Windows 10 or newer, Home or Pro
+* NVIDIA 452.39 or newer Drivers if you have an NVIDIA card
+* AMD Radeon Driver https://www.amd.com/en/support if you have a Radeon card
+
+## API Access
+
+Here's a quick example showing API access from `powershell`
+```powershell
+(Invoke-WebRequest -method POST -Body '{"model":"llama3", "prompt":"Why is the sky blue?", "stream": false}' -uri http://localhost:11434/api/generate ).Content | ConvertFrom-json
+```
+
+## Troubleshooting
+
+While we're in preview, `OLLAMA_DEBUG` is always enabled, which adds
+a "view logs" menu item to the app, and increases logging for the GUI app and
+server.
+
+Ollama on Windows stores files in a few different locations.  You can view them in
+the explorer window by hitting `<cmd>+R` and type in:
+- `explorer %LOCALAPPDATA%\Ollama` contains logs, and downloaded updates
+    - *app.log* contains logs from the GUI application
+    - *server.log* contains the server logs
+    - *upgrade.log* contains log output for upgrades
+- `explorer %LOCALAPPDATA%\Programs\Ollama` contains the binaries (The installer adds this to your user PATH)
+- `explorer %HOMEPATH%\.ollama` contains models and configuration
+- `explorer %TEMP%` contains temporary executable files in one or more `ollama*` directories
+
+
+## Standalone CLI
+
+The easiest way to install Ollama on Windows is to use the `OllamaSetup.exe`
+installer. It installs in your account without requiring Administrator rights.
+We update Ollama regularly to support the latest models, and this installer will
+help you keep up to date.
+
+If you'd like to install or integrate Ollama as a service, a standalone
+`ollama-windows-amd64.zip` zip file is available containing only the Ollama CLI
+and GPU library dependencies for Nvidia and AMD. This allows for embedding
+Ollama in existing applications, or running it as a system service via `ollama
+serve` with tools such as [NSSM](https://nssm.cc/).