Skip to content

GitLab

Unverified Commit df5e9c53 authored by YeAnbang's avatar YeAnbang Committed by GitHub
Browse files

[ColossalChat] Update RLHF V2 (#5286) 



* Add dpo. Fix sft, ppo, lora. Refactor all

* fix and tested ppo

* 2 nd round refactor

* add ci tests

* fix ci

* fix ci

* fix readme, style

* fix readme style

* fix style, fix benchmark

* reproduce benchmark result, remove useless files

* rename to ColossalChat

* use new image

* fix ci workflow

* fix ci

* use local model/tokenizer for ci tests

* fix ci

* fix ci

* fix ci

* fix ci timeout

* fix rm progress bar. fix ci timeout

* fix ci

* fix ci typo

* remove 3d plugin from ci temporary

* test environment

* cannot save optimizer

* support chat template

* fix readme

* fix path

* test ci locally

* restore build_or_pr

* fix ci data path

* fix benchmark

* fix ci, move ci tests to 3080, disable fast tokenizer

* move ci to 85

* support flash attention 2

* add all-in-one data preparation script. Fix colossal-llama2-chat chat template

* add hardware requirements

* move ci test data

* fix save_model, add unwrap

* fix missing bos

* fix missing bos; support grad accumulation with gemini

* fix ci

* fix ci

* fix ci

* fix llama2 chat template config

* debug sft

* debug sft

* fix colossalai version requirement

* fix ci

* add sanity check to prevent NaN loss

* fix requirements

* add dummy data generation script

* add dummy data generation script

* add dummy data generation script

* add dummy data generation script

* update readme

* update readme

* update readme and ignore

* fix logger bug

* support parallel_output

* modify data preparation logic

* fix tokenization

* update lr

* fix inference

* run pre-commit

---------
Co-authored-by: default avatarTong Li <tong.li352711588@gmail.com>
parent 36c4bb28
Expand all Hide whitespace changes
Inline Side-by-side
Showing with 0 additions and 2510 deletions
applications/Chat/coati/models/bloom/bloom_actor.py deleted 100644 → 0
View file @ 36c4bb28
from typing import Optional
from transformers import BloomConfig, BloomForCausalLM
from ..base import Actor
class BLOOMActor(Actor):
"""
BLOOM Actor model.
Args:
pretrained (str): Pretrained model name or path.
config (BloomConfig): Model config.
checkpoint (bool): Enable gradient checkpointing.
lora_rank (int): LoRA rank.
lora_train_bias (str): LoRA bias training mode.
"""
def __init__(
self,
pretrained: str = None,
config: Optional[BloomConfig] = None,
checkpoint: bool = False,
lora_rank: int = 0,
lora_train_bias: str = "none",
) -> None:
if pretrained is not None:
model = BloomForCausalLM.from_pretrained(pretrained)
elif config is not None:
model = BloomForCausalLM(config)
else:
model = BloomForCausalLM(BloomConfig())
if checkpoint:
model.gradient_checkpointing_enable()
super().__init__(model, lora_rank, lora_train_bias)
applications/Chat/coati/models/bloom/bloom_critic.py deleted 100644 → 0
View file @ 36c4bb28
from typing import Optional
import torch.nn as nn
from transformers import BloomConfig, BloomModel
from ..base import Critic
class BLOOMCritic(Critic):
"""
BLOOM Critic model.
Args:
pretrained (str): Pretrained model name or path.
config (BloomConfig): Model config.
lora_rank (int): LoRA rank.
lora_train_bias (str): LoRA bias training mode.
"""
def __init__(
self,
pretrained: str = None,
config: Optional[BloomConfig] = None,
lora_rank: int = 0,
lora_train_bias: str = "none",
**kwargs,
) -> None:
if pretrained is not None:
model = BloomModel.from_pretrained(pretrained)
elif config is not None:
model = BloomModel(config)
else:
model = BloomModel(BloomConfig())
value_head = nn.Linear(model.config.hidden_size, 1)
super().__init__(model, value_head, lora_rank, lora_train_bias, **kwargs)
applications/Chat/coati/models/bloom/bloom_rm.py deleted 100644 → 0
View file @ 36c4bb28
from typing import Optional
import torch.nn as nn
from transformers import BloomConfig, BloomModel
from ..base import RewardModel
class BLOOMRM(RewardModel):
"""
BLOOM Reward model.
Args:
pretrained (str): Pretrained model name or path.
config (BloomConfig): Model config.
lora_rank (int): LoRA rank.
lora_train_bias (str): LoRA bias training mode.
"""
def __init__(
self,
pretrained: str = None,
config: Optional[BloomConfig] = None,
lora_rank: int = 0,
lora_train_bias: str = "none",
) -> None:
if pretrained is not None:
model = BloomModel.from_pretrained(pretrained)
elif config is not None:
model = BloomModel(config)
else:
model = BloomModel(BloomConfig())
value_head = nn.Linear(model.config.hidden_size, 1)
value_head.weight.data.normal_(mean=0.0, std=1 / (model.config.hidden_size + 1))
super().__init__(model, value_head, lora_rank, lora_train_bias)
applications/Chat/coati/models/chatglm/__init__.py deleted 100644 → 0
View file @ 36c4bb28
from .chatglm_actor import ChatGLMActor
__all__ = ["ChatGLMActor"]
applications/Chat/coati/models/chatglm/chatglm_actor.py deleted 100644 → 0
View file @ 36c4bb28
from typing import Optional
from ..base import Actor
from .configuration_chatglm import ChatGLMConfig
from .modeling_chatglm import ChatGLMForConditionalGeneration
class ChatGLMActor(Actor):
"""
ChatGLM Actor model.
Args:
pretrained (str): Pretrained model name or path.
config (ChatGLMConfig): Model config.
checkpoint (bool): Enable gradient checkpointing.
do not support lora for now.
"""
def __init__(
self, pretrained: str = None, config: Optional[ChatGLMConfig] = None, checkpoint: bool = False
) -> None:
if pretrained is not None:
model = ChatGLMForConditionalGeneration.from_pretrained(pretrained)
elif config is not None:
model = ChatGLMForConditionalGeneration(config)
else:
model = ChatGLMForConditionalGeneration(ChatGLMConfig())
if checkpoint:
model.gradient_checkpointing_enable()
super().__init__(model, lora_rank=0, lora_train_bias="none")
applications/Chat/coati/models/chatglm/chatglm_tokenizer.py deleted 100644 → 0
View file @ 36c4bb28
"""
This code is copied from https://huggingface.co/THUDM/chatglm-6b/blob/main/tokenization_chatglm.py
"""
"""Tokenization classes for ChatGLM."""
import os
from typing import Dict, List, Optional, Union
import numpy as np
import sentencepiece as spm
from transformers.tokenization_utils import PreTrainedTokenizer
from transformers.tokenization_utils_base import BatchEncoding, EncodedInput
from transformers.utils import PaddingStrategy, logging
logger = logging.get_logger(__name__)
PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {
"THUDM/chatglm-6b": 2048,
}
class TextTokenizer:
def __init__(self, model_path):
self.sp = spm.SentencePieceProcessor()
self.sp.Load(model_path)
self.num_tokens = self.sp.vocab_size()
def encode(self, text):
return self.sp.EncodeAsIds(text)
def decode(self, ids: List[int]):
return self.sp.DecodeIds(ids)
def tokenize(self, text):
return self.sp.EncodeAsPieces(text)
def convert_tokens_to_string(self, tokens):
return self.sp.DecodePieces(tokens)
def convert_tokens_to_ids(self, tokens):
return [self.sp.PieceToId(token) for token in tokens]
def convert_token_to_id(self, token):
return self.sp.PieceToId(token)
def convert_id_to_token(self, idx):
return self.sp.IdToPiece(idx)
def __len__(self):
return self.num_tokens
class SPTokenizer:
def __init__(
self,
vocab_file,
num_image_tokens=20000,
max_blank_length=80,
byte_fallback=True,
):
assert vocab_file is not None
self.vocab_file = vocab_file
self.num_image_tokens = num_image_tokens
self.special_tokens = ["[MASK]", "[gMASK]", "[sMASK]", "<unused_0>", "<sop>", "<eop>", "<ENC>", "<dBLOCK>"]
self.max_blank_length = max_blank_length
self.byte_fallback = byte_fallback
self.text_tokenizer = TextTokenizer(vocab_file)
def _get_text_tokenizer(self):
return self.text_tokenizer
@staticmethod
def get_blank_token(length: int):
assert length >= 2
return f"<|blank_{length}|>"
@staticmethod
def get_tab_token():
return f"<|tab|>"
@property
def num_text_tokens(self):
return self.text_tokenizer.num_tokens
@property
def num_tokens(self):
return self.num_image_tokens + self.num_text_tokens
@staticmethod
def _encode_whitespaces(text: str, max_len: int = 80):
text = text.replace("\t", SPTokenizer.get_tab_token())
for i in range(max_len, 1, -1):
text = text.replace(" " * i, SPTokenizer.get_blank_token(i))
return text
def _preprocess(self, text: str, linebreak=True, whitespaces=True):
if linebreak:
text = text.replace("\n", "<n>")
if whitespaces:
text = self._encode_whitespaces(text, max_len=self.max_blank_length)
return text
def encode(self, text: str, linebreak=True, whitespaces=True, add_dummy_prefix=True) -> List[int]:
"""
@param text: Text to encode.
@param linebreak: Whether to encode newline (\n) in text.
@param whitespaces: Whether to encode multiple whitespaces or tab in text, useful for source code encoding.
@param special_tokens: Whether to encode special token ([MASK], [gMASK], etc.) in text.
@param add_dummy_prefix: Whether to add dummy blank space in the beginning.
"""
text = self._preprocess(text, linebreak, whitespaces)
if not add_dummy_prefix:
text = "<n>" + text
tmp = self._get_text_tokenizer().encode(text)
tokens = [x + self.num_image_tokens for x in tmp]
return tokens if add_dummy_prefix else tokens[2:]
def postprocess(self, text):
text = text.replace("<n>", "\n")
text = text.replace(SPTokenizer.get_tab_token(), "\t")
for i in range(2, self.max_blank_length + 1):
text = text.replace(self.get_blank_token(i), " " * i)
return text
def decode(self, text_ids: List[int]) -> str:
ids = [int(_id) - self.num_image_tokens for _id in text_ids]
ids = [_id for _id in ids if _id >= 0]
text = self._get_text_tokenizer().decode(ids)
text = self.postprocess(text)
return text
def decode_tokens(self, tokens: List[str]) -> str:
text = self._get_text_tokenizer().convert_tokens_to_string(tokens)
text = self.postprocess(text)
return text
def tokenize(self, text: str, linebreak=True, whitespaces=True, add_dummy_prefix=True) -> List[str]:
"""
@param text: Text to encode.
@param linebreak: Whether to encode newline (\n) in text.
@param whitespaces: Whether to encode multiple whitespaces or tab in text, useful for source code encoding.
@param special_tokens: Whether to encode special token ([MASK], [gMASK], etc.) in text.
@param add_dummy_prefix: Whether to add dummy blank space in the beginning.
"""
text = self._preprocess(text, linebreak, whitespaces)
if not add_dummy_prefix:
text = "<n>" + text
tokens = self._get_text_tokenizer().tokenize(text)
return tokens if add_dummy_prefix else tokens[2:]
def __getitem__(self, x: Union[int, str]):
if isinstance(x, int):
if x < self.num_image_tokens:
return "<image_{}>".format(x)
else:
return self.text_tokenizer.convert_id_to_token(x - self.num_image_tokens)
elif isinstance(x, str):
if x.startswith("<image_") and x.endswith(">") and x[7:-1].isdigit():
return int(x[7:-1])
else:
return self.text_tokenizer.convert_token_to_id(x) + self.num_image_tokens
else:
raise ValueError("The key should be str or int.")
class ChatGLMTokenizer(PreTrainedTokenizer):
"""
Construct a ChatGLM tokenizer. Based on byte-level Byte-Pair-Encoding.
Args:
vocab_file (`str`):
Path to the vocabulary file.
"""
vocab_files_names = {"vocab_file": "ice_text.model"}
max_model_input_sizes = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
model_input_names = ["input_ids", "attention_mask", "position_ids"]
def __init__(
self,
vocab_file,
do_lower_case=False,
remove_space=False,
bos_token="<sop>",
eos_token="<eop>",
end_token="</s>",
mask_token="[MASK]",
gmask_token="[gMASK]",
padding_side="left",
pad_token="<pad>",
unk_token="<unk>",
num_image_tokens=20000,
**kwargs,
) -> None:
super().__init__(
do_lower_case=do_lower_case,
remove_space=remove_space,
padding_side=padding_side,
bos_token=bos_token,
eos_token=eos_token,
end_token=end_token,
mask_token=mask_token,
gmask_token=gmask_token,
pad_token=pad_token,
unk_token=unk_token,
num_image_tokens=num_image_tokens,
**kwargs,
)
self.do_lower_case = do_lower_case
self.remove_space = remove_space
self.vocab_file = vocab_file
self.bos_token = bos_token
self.eos_token = eos_token
self.end_token = end_token
self.mask_token = mask_token
self.gmask_token = gmask_token
self.sp_tokenizer = SPTokenizer(vocab_file, num_image_tokens=num_image_tokens)
""" Initialisation """
@property
def gmask_token_id(self) -> Optional[int]:
if self.gmask_token is None:
return None
return self.convert_tokens_to_ids(self.gmask_token)
@property
def end_token_id(self) -> Optional[int]:
"""
`Optional[int]`: Id of the end of context token in the vocabulary. Returns `None` if the token has not been
set.
"""
if self.end_token is None:
return None
return self.convert_tokens_to_ids(self.end_token)
@property
def vocab_size(self):
"""Returns vocab size"""
return self.sp_tokenizer.num_tokens
def get_vocab(self):
"""Returns vocab as a dict"""
vocab = {self._convert_id_to_token(i): i for i in range(self.vocab_size)}
vocab.update(self.added_tokens_encoder)
return vocab
def preprocess_text(self, inputs):
if self.remove_space:
outputs = " ".join(inputs.strip().split())
else:
outputs = inputs
if self.do_lower_case:
outputs = outputs.lower()
return outputs
def _tokenize(self, text, **kwargs):
"""Returns a tokenized string."""
text = self.preprocess_text(text)
seq = self.sp_tokenizer.tokenize(text)
return seq
def convert_tokens_to_string(self, tokens: List[str]) -> str:
return self.sp_tokenizer.decode_tokens(tokens)
def _decode(self, token_ids: Union[int, List[int]], **kwargs) -> str:
if isinstance(token_ids, int):
token_ids = [token_ids]
if len(token_ids) == 0:
return ""
if self.pad_token_id in token_ids: # remove pad
token_ids = list(filter((self.pad_token_id).__ne__, token_ids))
return super()._decode(token_ids, **kwargs)
def _convert_token_to_id(self, token):
"""Converts a token (str) in an id using the vocab."""
return self.sp_tokenizer[token]
def _convert_id_to_token(self, index):
"""Converts an index (integer) in a token (str) using the vocab."""
return self.sp_tokenizer[index]
def save_vocabulary(self, save_directory, filename_prefix=None):
"""
Save the vocabulary and special tokens file to a directory.
Args:
save_directory (`str`):
The directory in which to save the vocabulary.
filename_prefix (`str`, *optional*):
An optional prefix to add to the named of the saved files.
Returns:
`Tuple(str)`: Paths to the files saved.
"""
if os.path.isdir(save_directory):
vocab_file = os.path.join(save_directory, self.vocab_files_names["vocab_file"])
else:
vocab_file = save_directory
with open(self.vocab_file, "rb") as fin:
proto_str = fin.read()
with open(vocab_file, "wb") as writer:
writer.write(proto_str)
return (vocab_file,)
def build_inputs_with_special_tokens(
self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
) -> List[int]:
"""
Build model inputs from a sequence or a pair of sequence for sequence classification tasks by concatenating and
adding special tokens. A BERT sequence has the following format:
- single sequence: `[CLS] X [SEP]`
- pair of sequences: `[CLS] A [SEP] B [SEP]`
Args:
token_ids_0 (`List[int]`):
List of IDs to which the special tokens will be added.
token_ids_1 (`List[int]`, *optional*):
Optional second list of IDs for sequence pairs.
Returns:
`List[int]`: List of [input IDs](../glossary#input-ids) with the appropriate special tokens.
"""
gmask_id = self.sp_tokenizer[self.gmask_token]
self.sp_tokenizer[self.eos_token]
token_ids_0 = token_ids_0 + [gmask_id, self.sp_tokenizer[self.bos_token]]
if token_ids_1 is not None:
token_ids_0 = token_ids_0 + token_ids_1
return token_ids_0
def _pad(
self,
encoded_inputs: Union[Dict[str, EncodedInput], BatchEncoding],
max_length: Optional[int] = None,
padding_strategy: PaddingStrategy = PaddingStrategy.DO_NOT_PAD,
pad_to_multiple_of: Optional[int] = None,
return_attention_mask: Optional[bool] = None,
) -> dict:
"""
Pad encoded inputs (on left/right and up to predefined length or max length in the batch)
Args:
encoded_inputs:
Dictionary of tokenized inputs (`List[int]`) or batch of tokenized inputs (`List[List[int]]`).
max_length: maximum length of the returned list and optionally padding length (see below).
Will truncate by taking into account the special tokens.
padding_strategy: PaddingStrategy to use for padding.
- PaddingStrategy.LONGEST Pad to the longest sequence in the batch
- PaddingStrategy.MAX_LENGTH: Pad to the max length (default)
- PaddingStrategy.DO_NOT_PAD: Do not pad
The tokenizer padding sides are defined in self.padding_side:
- 'left': pads on the left of the sequences
- 'right': pads on the right of the sequences
pad_to_multiple_of: (optional) Integer if set will pad the sequence to a multiple of the provided value.
This is especially useful to enable the use of Tensor Core on NVIDIA hardware with compute capability
`>= 7.5` (Volta).
return_attention_mask:
(optional) Set to False to avoid returning attention mask (default: set to model specifics)
"""
# Load from model defaults
bos_token_id = self.sp_tokenizer[self.bos_token]
mask_token_id = self.sp_tokenizer[self.mask_token]
gmask_token_id = self.sp_tokenizer[self.gmask_token]
assert self.padding_side == "left"
required_input = encoded_inputs[self.model_input_names[0]]
seq_length = len(required_input)
if padding_strategy == PaddingStrategy.LONGEST:
max_length = len(required_input)
if max_length is not None and pad_to_multiple_of is not None and (max_length % pad_to_multiple_of != 0):
max_length = ((max_length // pad_to_multiple_of) + 1) * pad_to_multiple_of
needs_to_be_padded = padding_strategy != PaddingStrategy.DO_NOT_PAD and len(required_input) != max_length
# Initialize attention mask if not present.
if max_length is not None:
if "attention_mask" not in encoded_inputs:
if bos_token_id in required_input:
context_length = required_input.index(bos_token_id)
else:
context_length = seq_length
attention_mask = np.ones((1, seq_length, seq_length))
attention_mask = np.tril(attention_mask)
attention_mask[:, :, :context_length] = 1
attention_mask = np.bool_(attention_mask < 0.5)
encoded_inputs["attention_mask"] = attention_mask
if "position_ids" not in encoded_inputs:
if bos_token_id in required_input:
context_length = required_input.index(bos_token_id)
else:
context_length = seq_length
position_ids = np.arange(seq_length, dtype=np.int64)
mask_token = mask_token_id if mask_token_id in required_input else gmask_token_id
if mask_token in required_input:
mask_position = required_input.index(mask_token)
position_ids[context_length:] = mask_position
block_position_ids = np.concatenate(
[
np.zeros(context_length, dtype=np.int64),
np.arange(1, seq_length - context_length + 1, dtype=np.int64),
]
)
encoded_inputs["position_ids"] = np.stack([position_ids, block_position_ids], axis=0)
if needs_to_be_padded:
difference = max_length - len(required_input)
if "attention_mask" in encoded_inputs:
encoded_inputs["attention_mask"] = np.pad(
encoded_inputs["attention_mask"],
pad_width=[(0, 0), (difference, 0), (difference, 0)],
mode="constant",
constant_values=True,
)
if "token_type_ids" in encoded_inputs:
encoded_inputs["token_type_ids"] = [self.pad_token_type_id] * difference + encoded_inputs[
"token_type_ids"
]
if "special_tokens_mask" in encoded_inputs:
encoded_inputs["special_tokens_mask"] = [1] * difference + encoded_inputs["special_tokens_mask"]
if "position_ids" in encoded_inputs:
encoded_inputs["position_ids"] = np.pad(
encoded_inputs["position_ids"], pad_width=[(0, 0), (difference, 0)]
)
encoded_inputs[self.model_input_names[0]] = [self.pad_token_id] * difference + required_input
return encoded_inputs
applications/Chat/coati/models/chatglm/configuration_chatglm.py deleted 100644 → 0
View file @ 36c4bb28
"""
This code is copied from https://huggingface.co/THUDM/chatglm-6b/resolve/main/configuration_chatglm.py
"""
""" ChatGLM model configuration """
from transformers.configuration_utils import PretrainedConfig
from transformers.utils import logging
logger = logging.get_logger(__name__)
class ChatGLMConfig(PretrainedConfig):
r"""
This is the configuration class to store the configuration of a [`~ChatGLMModel`].
It is used to instantiate an ChatGLM model according to the specified arguments, defining the model
architecture. Instantiating a configuration with the defaults will yield a similar configuration to that of
the ChatGLM-6B [THUDM/ChatGLM-6B](https://huggingface.co/THUDM/chatglm-6b) architecture.
Configuration objects inherit from [`PretrainedConfig`] and can be used
to control the model outputs. Read the documentation from [`PretrainedConfig`]
for more information.
Args:
vocab_size (`int`, *optional*, defaults to 150528):
Vocabulary size of the ChatGLM-6B model. Defines the number of different tokens that can be represented by the
`inputs_ids` passed when calling [`~ChatGLMModel`] or
[`~TFChatGLMModel`].
hidden_size (`int`, *optional*, defaults to 4096):
Dimension of the encoder layers and the pooler layer.
num_hidden_layers (`int`, *optional*, defaults to 28):
Number of hidden layers in the Transformer encoder.
num_attention_heads (`int`, *optional*, defaults to 32):
Number of attention heads for each attention layer in the Transformer encoder.
inner_hidden_size (`int`, *optional*, defaults to 16384):
Dimension of the "intermediate" (i.e., feed-forward) layer in the Transformer encoder.
max_sequence_length (`int`, *optional*, defaults to 512):
The maximum sequence length that this model might ever be used with.
Typically set this to something large just in case (e.g., 512 or 1024 or 2048).
layernorm_epsilon (`float`, *optional*, defaults to 1e-5):
The epsilon used by the layer normalization layers.
use_cache (`bool`, *optional*, defaults to `True`):
Whether the model should return the last key/values attentions (not used by all models).
Example:
```python
>>> from configuration_chatglm import ChatGLMConfig
>>> from modeling_chatglm import ChatGLMModel
>>> # Initializing a ChatGLM-6B THUDM/ChatGLM-6B style configuration
>>> configuration = ChatGLMConfig()
>>> # Initializing a model from the THUDM/ChatGLM-6B style configuration
>>> model = ChatGLMModel(configuration)
>>> # Accessing the model configuration
>>> configuration = model.config
```"""
model_type = "chatglm"
def __init__(
self,
vocab_size=130528,
hidden_size=4096,
num_layers=28,
num_attention_heads=32,
layernorm_epsilon=1e-5,
use_cache=True,
bos_token_id=130004,
eos_token_id=130005,
mask_token_id=130000,
gmask_token_id=130001,
pad_token_id=3,
max_sequence_length=2048,
inner_hidden_size=16384,
position_encoding_2d=True,
quantization_bit=0,
pre_seq_len=None,
prefix_projection=False,
**kwargs,
):
self.num_layers = num_layers
self.vocab_size = vocab_size
self.hidden_size = hidden_size
self.num_attention_heads = num_attention_heads
self.max_sequence_length = max_sequence_length
self.layernorm_epsilon = layernorm_epsilon
self.inner_hidden_size = inner_hidden_size
self.use_cache = use_cache
self.bos_token_id = bos_token_id
self.eos_token_id = eos_token_id
self.pad_token_id = pad_token_id
self.mask_token_id = mask_token_id
self.gmask_token_id = gmask_token_id
self.position_encoding_2d = position_encoding_2d
self.quantization_bit = quantization_bit
self.pre_seq_len = pre_seq_len
self.prefix_projection = prefix_projection
super().__init__(pad_token_id=pad_token_id, bos_token_id=bos_token_id, eos_token_id=eos_token_id, **kwargs)
applications/Chat/coati/models/gpt/__init__.py deleted 100644 → 0
View file @ 36c4bb28
from .gpt_actor import GPTActor
from .gpt_critic import GPTCritic
from .gpt_rm import GPTRM
__all__ = ["GPTActor", "GPTCritic", "GPTRM"]
applications/Chat/coati/models/gpt/gpt_actor.py deleted 100644 → 0
View file @ 36c4bb28
from typing import Optional
from transformers.models.gpt2.configuration_gpt2 import GPT2Config
from transformers.models.gpt2.modeling_gpt2 import GPT2LMHeadModel
from ..base import Actor
class GPTActor(Actor):
"""
GPT Actor model.
Args:
pretrained (str): Pretrained model name or path.
config (GPT2Config): Model config.
checkpoint (bool): Enable gradient checkpointing.
lora_rank (int): Rank of the LoRa layer.
lora_train_bias (str): Bias training strategy for the LoRa layer.
"""
def __init__(
self,
pretrained: Optional[str] = None,
config: Optional[GPT2Config] = None,
checkpoint: bool = False,
lora_rank: int = 0,
lora_train_bias: str = "none",
**kwargs,
) -> None:
if pretrained is not None:
model = GPT2LMHeadModel.from_pretrained(pretrained)
elif config is not None:
model = GPT2LMHeadModel(config)
else:
model = GPT2LMHeadModel(GPT2Config())
if checkpoint:
model.gradient_checkpointing_enable()
super().__init__(model, lora_rank, lora_train_bias, **kwargs)
applications/Chat/coati/models/gpt/gpt_critic.py deleted 100644 → 0
View file @ 36c4bb28
from typing import Optional
import torch.nn as nn
from transformers.models.gpt2.configuration_gpt2 import GPT2Config
from transformers.models.gpt2.modeling_gpt2 import GPT2Model
from ..base import Critic
class GPTCritic(Critic):
"""
GPT Critic model.
Args:
pretrained (str): Pretrained model name or path.
config (GPT2Config): Model config.
lora_rank (int): Rank of the LO-RA decomposition.
lora_train_bias (str): LoRA bias training mode.
"""
def __init__(
self,
pretrained: Optional[str] = None,
config: Optional[GPT2Config] = None,
lora_rank: int = 0,
lora_train_bias: str = "none",
**kwargs,
) -> None:
if pretrained is not None:
model = GPT2Model.from_pretrained(pretrained)
elif config is not None:
model = GPT2Model(config)
else:
model = GPT2Model(GPT2Config())
value_head = nn.Linear(model.config.n_embd, 1)
super().__init__(model, value_head, lora_rank, lora_train_bias, **kwargs)
applications/Chat/coati/models/gpt/gpt_rm.py deleted 100644 → 0
View file @ 36c4bb28
from typing import Optional
import torch.nn as nn
from transformers.models.gpt2.configuration_gpt2 import GPT2Config
from transformers.models.gpt2.modeling_gpt2 import GPT2Model
from ..base import RewardModel
class GPTRM(RewardModel):
"""
GPT Reward model.
Args:
pretrained (str): Pretrained model name or path.
config (GPT2Config): Model config.
lora_rank (int): Rank of the low-rank approximation.
lora_train_bias (str): LoRA bias training mode.
"""
def __init__(
self,
pretrained: Optional[str] = None,
config: Optional[GPT2Config] = None,
lora_rank: int = 0,
lora_train_bias: str = "none",
) -> None:
if pretrained is not None:
model = GPT2Model.from_pretrained(pretrained)
elif config is not None:
model = GPT2Model(config)
else:
model = GPT2Model(GPT2Config())
value_head = nn.Linear(model.config.n_embd, 1)
value_head.weight.data.normal_(mean=0.0, std=1 / (model.config.n_embd + 1))
super().__init__(model, value_head, lora_rank, lora_train_bias)
applications/Chat/coati/models/llama/__init__.py deleted 100644 → 0
View file @ 36c4bb28
from .llama_actor import LlamaActor
from .llama_critic import LlamaCritic
from .llama_rm import LlamaRM
__all__ = ["LlamaActor", "LlamaCritic", "LlamaRM"]
applications/Chat/coati/models/llama/llama_actor.py deleted 100644 → 0
View file @ 36c4bb28
from typing import Optional
from transformers import LlamaConfig, LlamaForCausalLM
from ..base import Actor
class LlamaActor(Actor):
"""
Llama Actor model.
Args:
pretrained (str): Pretrained model name or path.
config (LlamaConfig): Model config.
checkpoint (bool): Enable gradient checkpointing.
lora_rank (int): LoRA rank.
lora_train_bias (str): LoRA bias training mode.
"""
def __init__(
self,
pretrained: Optional[str] = None,
config: Optional[LlamaConfig] = None,
checkpoint: bool = False,
lora_rank: int = 0,
lora_train_bias: str = "none",
) -> None:
if pretrained is not None:
model = LlamaForCausalLM.from_pretrained(pretrained)
elif config is not None:
model = LlamaForCausalLM(config)
else:
model = LlamaForCausalLM(LlamaConfig())
if checkpoint:
model.gradient_checkpointing_enable()
super().__init__(model, lora_rank, lora_train_bias)
applications/Chat/coati/models/llama/llama_critic.py deleted 100644 → 0
View file @ 36c4bb28
from typing import Optional
import torch.nn as nn
from transformers import LlamaConfig, LlamaModel
from ..base import Critic
class LlamaCritic(Critic):
"""
Llama Critic model.
Args:
pretrained (str): Pretrained model name or path.
config (LlamaConfig): Model config.
lora_rank (int): LoRA rank.
lora_train_bias (str): LoRA bias training mode.
"""
def __init__(
self,
pretrained: Optional[str] = None,
config: Optional[LlamaConfig] = None,
lora_rank: int = 0,
lora_train_bias: str = "none",
**kwargs,
) -> None:
if pretrained is not None:
model = LlamaModel.from_pretrained(pretrained)
elif config is not None:
model = LlamaModel(config)
else:
model = LlamaModel(LlamaConfig())
value_head = nn.Linear(model.config.hidden_size, 1)
super().__init__(model, value_head, lora_rank, lora_train_bias, **kwargs)
applications/Chat/coati/models/llama/llama_rm.py deleted 100644 → 0
View file @ 36c4bb28
from typing import Optional
import torch.nn as nn
from transformers import LlamaConfig, LlamaModel
from ..base import RewardModel
class LlamaRM(RewardModel):
"""
Llama Reward model.
Args:
pretrained (str): Pretrained model name or path.
config (LlamaConfig): Model config.
lora_rank (int): LoRA rank.
lora_train_bias (str): LoRA bias training mode.
"""
def __init__(
self,
pretrained: Optional[str] = None,
config: Optional[LlamaConfig] = None,
lora_rank: int = 0,
lora_train_bias: str = "none",
) -> None:
if pretrained is not None:
model = LlamaModel.from_pretrained(pretrained)
elif config is not None:
model = LlamaModel(config)
else:
model = LlamaModel(LlamaConfig())
value_head = nn.Linear(model.config.hidden_size, 1)
value_head.weight.data.normal_(mean=0.0, std=1 / (model.config.hidden_size + 1))
super().__init__(model, value_head, lora_rank, lora_train_bias)
applications/Chat/coati/models/opt/__init__.py deleted 100644 → 0
View file @ 36c4bb28
from .opt_actor import OPTActor
from .opt_critic import OPTCritic
from .opt_rm import OPTRM
__all__ = ["OPTActor", "OPTCritic", "OPTRM"]
applications/Chat/coati/models/opt/opt_actor.py deleted 100644 → 0
View file @ 36c4bb28
from typing import Optional
from transformers.models.opt.configuration_opt import OPTConfig
from transformers.models.opt.modeling_opt import OPTForCausalLM
from ..base import Actor
class OPTActor(Actor):
"""
OPT Actor model.
Args:
pretrained (str): Pretrained model name or path.
config (OPTConfig): Model config.
checkpoint (bool): Enable gradient checkpointing.
lora_rank (int): Rank of the low-rank approximation.
lora_train_bias (str): LoRA bias training mode.
"""
def __init__(
self,
pretrained: Optional[str] = None,
config: Optional[OPTConfig] = None,
checkpoint: bool = False,
lora_rank: int = 0,
lora_train_bias: str = "none",
) -> None:
if pretrained is not None:
model = OPTForCausalLM.from_pretrained(pretrained)
elif config is not None:
model = OPTForCausalLM(config)
else:
model = OPTForCausalLM(OPTConfig())
if checkpoint:
model.gradient_checkpointing_enable()
super().__init__(model, lora_rank, lora_train_bias)
applications/Chat/coati/models/opt/opt_critic.py deleted 100644 → 0
View file @ 36c4bb28
from typing import Optional
import torch.nn as nn
from transformers.models.opt.configuration_opt import OPTConfig
from transformers.models.opt.modeling_opt import OPTModel
from ..base import Critic
class OPTCritic(Critic):
"""
OPT Critic model.
Args:
pretrained (str): Pretrained model name or path.
config (OPTConfig): Model config.
lora_rank (int): Rank of the low-rank approximation.
lora_train_bias (str): LoRA bias training mode.
"""
def __init__(
self,
pretrained: Optional[str] = None,
config: Optional[OPTConfig] = None,
lora_rank: int = 0,
lora_train_bias: str = "none",
**kwargs,
) -> None:
if pretrained is not None:
model = OPTModel.from_pretrained(pretrained)
elif config is not None:
model = OPTModel(config)
else:
model = OPTModel(OPTConfig())
value_head = nn.Linear(model.config.word_embed_proj_dim, 1)
super().__init__(model, value_head, lora_rank, lora_train_bias, **kwargs)
applications/Chat/coati/models/opt/opt_rm.py deleted 100644 → 0
View file @ 36c4bb28
from typing import Optional
import torch.nn as nn
from transformers import OPTConfig, OPTModel
from ..base import RewardModel
class OPTRM(RewardModel):
"""
OPT Reward model.
Args:
pretrained (str): Pretrained model name or path.
config (OPTConfig): Model config.
lora_rank (int): Rank of the low-rank approximation.
lora_train_bias (str): LoRA bias training mode.
"""
def __init__(
self,
pretrained: Optional[str] = None,
config: Optional[OPTConfig] = None,
lora_rank: int = 0,
lora_train_bias: str = "none",
) -> None:
if pretrained is not None:
model = OPTModel.from_pretrained(pretrained)
elif config is not None:
model = OPTModel(config)
else:
model = OPTModel(OPTConfig())
value_head = nn.Linear(model.config.word_embed_proj_dim, 1)
value_head.weight.data.normal_(mean=0.0, std=1 / (model.config.word_embed_proj_dim + 1))
super().__init__(model, value_head, lora_rank, lora_train_bias)
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment