Commit 0211193c authored by zhuwenwen's avatar zhuwenwen
Browse files

initial llama

parents
Pipeline #509 failed with stages
in 0 seconds
Language: Cpp
AccessModifierOffset: -4
AlignAfterOpenBracket: Align
AllowShortEnumsOnASingleLine: false
AlignConsecutiveAssignments: true
AlignConsecutiveDeclarations: true
AlignEscapedNewlines: Right
AlignOperands: true
AlignTrailingComments: true
AllowAllParametersOfDeclarationOnNextLine: true
AllowAllArgumentsOnNextLine: true
AllowShortBlocksOnASingleLine: Empty
AllowShortCaseLabelsOnASingleLine: false
AllowShortFunctionsOnASingleLine: Empty
AllowShortIfStatementsOnASingleLine: Never
AllowShortLoopsOnASingleLine: false
AlwaysBreakAfterReturnType: None
AlwaysBreakBeforeMultilineStrings: false
AlwaysBreakTemplateDeclarations: true
BinPackArguments: false
BinPackParameters: false
BreakBeforeBinaryOperators: NonAssignment
BreakBeforeBraces: Stroustrup
BreakBeforeTernaryOperators: false
BreakConstructorInitializers: AfterColon
BreakInheritanceList: AfterColon
BreakStringLiterals: false
ColumnLimit: 120
CompactNamespaces: false
ConstructorInitializerAllOnOneLineOrOnePerLine: true
ConstructorInitializerIndentWidth: 4
ContinuationIndentWidth: 4
Cpp11BracedListStyle: true
DerivePointerAlignment: false
FixNamespaceComments: true
IndentCaseLabels: true
IndentPPDirectives: None
IndentWidth: 4
IndentWrappedFunctionNames: false
KeepEmptyLinesAtTheStartOfBlocks: true
MaxEmptyLinesToKeep: 1
NamespaceIndentation: None
PointerAlignment: Left
ReflowComments: true
SortIncludes: true
SortUsingDeclarations: false
SpaceAfterCStyleCast: false
SpaceAfterTemplateKeyword: false
SpaceBeforeAssignmentOperators: true
SpaceBeforeCtorInitializerColon: false
SpaceBeforeInheritanceColon: false
SpaceBeforeParens: ControlStatements
SpaceInEmptyParentheses: false
SpacesBeforeTrailingComments: 2
SpacesInAngles: false
SpacesInCStyleCastParentheses: false
SpacesInContainerLiterals: false
SpacesInParentheses: false
SpacesInSquareBrackets: false
Standard: Cpp11
TabWidth: 4
UseTab: Never
docker
.dockerignore
.gitlab
.gitlab-ci.yml
*build*
./models
__pycache__
.vscode
translation
.cache
*.npy
*.pth
*.o
**/.ipynb_checkpoints
\ No newline at end of file
[flake8]
ignore = W292
exclude =
*migrations*,
# python related
*.pyc,
.git,
__pycache__,
max-line-length=120
max-complexity=12
format=pylint
show_source = True
statistics = True
count = True
name: "Bug Report"
description: Submit a bug report
labels: [ "bug" ]
body:
- type: input
id: branch
attributes:
label: Branch/Tag/Commit
description:
placeholder: ex,. main
validations:
required: true
- type: input
id: docker_image_version
attributes:
label: Docker Image Version
description:
placeholder: ex,. nvcr.io/nvidia/pytorch:22.08-py3
validations:
required: true
- type: input
id: gpu_name
attributes:
label: GPU name
description:
placeholder: ex,. A100
validations:
required: true
- type: input
id: cuda_driver
attributes:
label: CUDA Driver
description:
placeholder: ex,. 515.65.01
validations:
required: true
- type: textarea
id: reproduced-steps
attributes:
label: Reproduced Steps
description: Please provide the step to reproduce the bugs
render: shell
placeholder: |
Steps to reproduce your bugs: (please list what script you run and don't say following xxx, otherwise, we will ask again and again)
1. docker run -ti --gpus all nvcr.io/nvidia/pytorch:22.03-py3 bash
2. git clone https://github.com/NVIDIA/FasterTransformer.git
3. cd FasterTransformer mkdir build && cd build
4. cmake -DSM=80 -DCMAKE_BUILD_TYPE=Release .. && make -j12
5. ./bin/bert_example 32 12 32 12 64 0 0
6. What error you see.
validations:
required: true
*~
*.o
*build*/
./models/
__pycache__/
.vscode
.idea
./translation
.cache
*.npy
*.pth
!tests/data/**/*.npy
/models
/notebooks
**/.ipynb_checkpoints/
.DS_Store
/3rdparty/NeMo/
/3rdparty/apex/
[submodule "3rdparty/Megatron-LM"]
path = 3rdparty/Megatron-LM
url = https://github.com/NVIDIA/Megatron-LM.git
branch = v2.6
[submodule "examples/tensorflow/bert/tensorflow_bert/bert"]
path = examples/tensorflow/bert/tensorflow_bert/bert
url = https://github.com/google-research/bert.git
[submodule "examples/pytorch/swin/Swin-Transformer-Quantization/SwinTransformer"]
path = examples/pytorch/swin/Swin-Transformer-Quantization/SwinTransformer
url = https://github.com/microsoft/Swin-Transformer
[submodule "examples/pytorch/vit/ViT-quantization/ViT-pytorch"]
path = examples/pytorch/vit/ViT-quantization/ViT-pytorch
url = https://github.com/jeonsworld/ViT-pytorch
[submodule "3rdparty/cutlass"]
path = 3rdparty/cutlass
url = https://github.com/NVIDIA/cutlass.git
# Copyright (c) 2019-2023, NVIDIA CORPORATION. All rights reserved.
#
# 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.
add_subdirectory(common)
add_subdirectory(trt_fused_multihead_attention)
if(ENABLE_FP8)
add_subdirectory(fp8_qgmma_1x1)
endif()
\ No newline at end of file
// Read an INI file into easy-to-access name/value pairs.
// inih and INIReader are released under the New BSD license.
// Go to the project home page for more info:
//
// https://github.com/benhoyt/inih (Initial repo)
// https://github.com/jtilly/inih (The reference of this header file)
/* inih -- simple .INI file parser
inih is released under the New BSD license (see LICENSE.txt). Go to the project
home page for more info:
https://github.com/benhoyt/inih
https://github.com/jtilly/inih
*/
#ifndef __INI_H__
#define __INI_H__
/* Make this header file easier to include in C++ code */
#ifdef __cplusplus
extern "C" {
#endif
#include <stdio.h>
/* Typedef for prototype of handler function. */
typedef int (*ini_handler)(void* user, const char* section,
const char* name, const char* value);
/* Typedef for prototype of fgets-style reader function. */
typedef char* (*ini_reader)(char* str, int num, void* stream);
/* Parse given INI-style file. May have [section]s, name=value pairs
(whitespace stripped), and comments starting with ';' (semicolon). Section
is "" if name=value pair parsed before any section heading. name:value
pairs are also supported as a concession to Python's configparser.
For each name=value pair parsed, call handler function with given user
pointer as well as section, name, and value (data only valid for duration
of handler call). Handler should return nonzero on success, zero on error.
Returns 0 on success, line number of first error on parse error (doesn't
stop on first error), -1 on file open error, or -2 on memory allocation
error (only when INI_USE_STACK is zero).
*/
int ini_parse(const char* filename, ini_handler handler, void* user);
/* Same as ini_parse(), but takes a FILE* instead of filename. This doesn't
close the file when it's finished -- the caller must do that. */
int ini_parse_file(FILE* file, ini_handler handler, void* user);
/* Same as ini_parse(), but takes an ini_reader function pointer instead of
filename. Used for implementing custom or string-based I/O. */
int ini_parse_stream(ini_reader reader, void* stream, ini_handler handler,
void* user);
/* Nonzero to allow multi-line value parsing, in the style of Python's
configparser. If allowed, ini_parse() will call the handler with the same
name for each subsequent line parsed. */
#ifndef INI_ALLOW_MULTILINE
#define INI_ALLOW_MULTILINE 1
#endif
/* Nonzero to allow a UTF-8 BOM sequence (0xEF 0xBB 0xBF) at the start of
the file. See http://code.google.com/p/inih/issues/detail?id=21 */
#ifndef INI_ALLOW_BOM
#define INI_ALLOW_BOM 1
#endif
/* Nonzero to allow inline comments (with valid inline comment characters
specified by INI_INLINE_COMMENT_PREFIXES). Set to 0 to turn off and match
Python 3.2+ configparser behaviour. */
#ifndef INI_ALLOW_INLINE_COMMENTS
#define INI_ALLOW_INLINE_COMMENTS 1
#endif
#ifndef INI_INLINE_COMMENT_PREFIXES
#define INI_INLINE_COMMENT_PREFIXES ";"
#endif
/* Nonzero to use stack, zero to use heap (malloc/free). */
#ifndef INI_USE_STACK
#define INI_USE_STACK 1
#endif
/* Stop parsing on first error (default is to keep parsing). */
#ifndef INI_STOP_ON_FIRST_ERROR
#define INI_STOP_ON_FIRST_ERROR 0
#endif
/* Maximum line length for any line in INI file. */
#ifndef INI_MAX_LINE
#define INI_MAX_LINE 200
#endif
#ifdef __cplusplus
}
#endif
/* inih -- simple .INI file parser
inih is released under the New BSD license (see LICENSE.txt). Go to the project
home page for more info:
https://github.com/benhoyt/inih
*/
#if defined(_MSC_VER) && !defined(_CRT_SECURE_NO_WARNINGS)
#define _CRT_SECURE_NO_WARNINGS
#endif
#include <stdio.h>
#include <ctype.h>
#include <string.h>
#if !INI_USE_STACK
#include <stdlib.h>
#endif
#define MAX_SECTION 50
#define MAX_NAME 50
/* Strip whitespace chars off end of given string, in place. Return s. */
inline static char* rstrip(char* s)
{
char* p = s + strlen(s);
while (p > s && isspace((unsigned char)(*--p)))
*p = '\0';
return s;
}
/* Return pointer to first non-whitespace char in given string. */
inline static char* lskip(const char* s)
{
while (*s && isspace((unsigned char)(*s)))
s++;
return (char*)s;
}
/* Return pointer to first char (of chars) or inline comment in given string,
or pointer to null at end of string if neither found. Inline comment must
be prefixed by a whitespace character to register as a comment. */
inline static char* find_chars_or_comment(const char* s, const char* chars)
{
#if INI_ALLOW_INLINE_COMMENTS
int was_space = 0;
while (*s && (!chars || !strchr(chars, *s)) &&
!(was_space && strchr(INI_INLINE_COMMENT_PREFIXES, *s))) {
was_space = isspace((unsigned char)(*s));
s++;
}
#else
while (*s && (!chars || !strchr(chars, *s))) {
s++;
}
#endif
return (char*)s;
}
/* Version of strncpy that ensures dest (size bytes) is null-terminated. */
inline static char* strncpy0(char* dest, const char* src, size_t size)
{
strncpy(dest, src, size);
dest[size - 1] = '\0';
return dest;
}
/* See documentation in header file. */
inline int ini_parse_stream(ini_reader reader, void* stream, ini_handler handler,
void* user)
{
/* Uses a fair bit of stack (use heap instead if you need to) */
#if INI_USE_STACK
char line[INI_MAX_LINE];
#else
char* line;
#endif
char section[MAX_SECTION] = "";
char prev_name[MAX_NAME] = "";
char* start;
char* end;
char* name;
char* value;
int lineno = 0;
int error = 0;
#if !INI_USE_STACK
line = (char*)malloc(INI_MAX_LINE);
if (!line) {
return -2;
}
#endif
/* Scan through stream line by line */
while (reader(line, INI_MAX_LINE, stream) != NULL) {
lineno++;
start = line;
#if INI_ALLOW_BOM
if (lineno == 1 && (unsigned char)start[0] == 0xEF &&
(unsigned char)start[1] == 0xBB &&
(unsigned char)start[2] == 0xBF) {
start += 3;
}
#endif
start = lskip(rstrip(start));
if (*start == ';' || *start == '#') {
/* Per Python configparser, allow both ; and # comments at the
start of a line */
}
#if INI_ALLOW_MULTILINE
else if (*prev_name && *start && start > line) {
#if INI_ALLOW_INLINE_COMMENTS
end = find_chars_or_comment(start, NULL);
if (*end)
*end = '\0';
rstrip(start);
#endif
/* Non-blank line with leading whitespace, treat as continuation
of previous name's value (as per Python configparser). */
if (!handler(user, section, prev_name, start) && !error)
error = lineno;
}
#endif
else if (*start == '[') {
/* A "[section]" line */
end = find_chars_or_comment(start + 1, "]");
if (*end == ']') {
*end = '\0';
strncpy0(section, start + 1, sizeof(section));
*prev_name = '\0';
}
else if (!error) {
/* No ']' found on section line */
error = lineno;
}
}
else if (*start) {
/* Not a comment, must be a name[=:]value pair */
end = find_chars_or_comment(start, "=:");
if (*end == '=' || *end == ':') {
*end = '\0';
name = rstrip(start);
value = lskip(end + 1);
#if INI_ALLOW_INLINE_COMMENTS
end = find_chars_or_comment(value, NULL);
if (*end)
*end = '\0';
#endif
rstrip(value);
/* Valid name[=:]value pair found, call handler */
strncpy0(prev_name, name, sizeof(prev_name));
if (!handler(user, section, name, value) && !error)
error = lineno;
}
else if (!error) {
/* No '=' or ':' found on name[=:]value line */
error = lineno;
}
}
#if INI_STOP_ON_FIRST_ERROR
if (error)
break;
#endif
}
#if !INI_USE_STACK
free(line);
#endif
return error;
}
/* See documentation in header file. */
inline int ini_parse_file(FILE* file, ini_handler handler, void* user)
{
return ini_parse_stream((ini_reader)fgets, file, handler, user);
}
/* See documentation in header file. */
inline int ini_parse(const char* filename, ini_handler handler, void* user)
{
FILE* file;
int error;
file = fopen(filename, "r");
if (!file)
return -1;
error = ini_parse_file(file, handler, user);
fclose(file);
return error;
}
#endif /* __INI_H__ */
#ifndef __INIREADER_H__
#define __INIREADER_H__
#include <map>
#include <set>
#include <string>
// Read an INI file into easy-to-access name/value pairs. (Note that I've gone
// for simplicity here rather than speed, but it should be pretty decent.)
class INIReader
{
public:
// Empty Constructor
INIReader() {};
// Construct INIReader and parse given filename. See ini.h for more info
// about the parsing.
INIReader(std::string filename);
// Construct INIReader and parse given file. See ini.h for more info
// about the parsing.
INIReader(FILE *file);
~INIReader();
// Return the result of ini_parse(), i.e., 0 on success, line number of
// first error on parse error, or -1 on file open error.
int ParseError() const;
// Return the list of sections found in ini file
const std::set<std::string>& Sections() const;
// Get a string value from INI file, returning default_value if not found.
std::string Get(std::string section, std::string name,
std::string default_value) const;
std::string Get(std::string section, std::string name) const;
// Get an integer (long) value from INI file, returning default_value if
// not found or not a valid integer (decimal "1234", "-1234", or hex "0x4d2").
long GetInteger(std::string section, std::string name, long default_value) const;
long GetInteger(std::string section, std::string name) const;
// Get a real (floating point double) value from INI file, returning
// default_value if not found or not a valid floating point value
// according to strtod().
double GetReal(std::string section, std::string name, double default_value) const;
// Get a single precision floating point number value from INI file, returning
// default_value if not found or not a valid floating point value
// according to strtof().
float GetFloat(std::string section, std::string name, float default_value) const;
float GetFloat(std::string section, std::string name) const;
// Get a boolean value from INI file, returning default_value if not found or if
// not a valid true/false value. Valid true values are "true", "yes", "on", "1",
// and valid false values are "false", "no", "off", "0" (not case sensitive).
bool GetBoolean(std::string section, std::string name, bool default_value) const;
protected:
int _error;
std::map<std::string, std::string> _values;
std::set<std::string> _sections;
static std::string MakeKey(std::string section, std::string name);
static int ValueHandler(void* user, const char* section, const char* name,
const char* value);
};
#endif // __INIREADER_H__
#ifndef __INIREADER__
#define __INIREADER__
#include <algorithm>
#include <cctype>
#include <cstdlib>
inline INIReader::INIReader(std::string filename)
{
_error = ini_parse(filename.c_str(), ValueHandler, this);
}
inline INIReader::INIReader(FILE *file)
{
_error = ini_parse_file(file, ValueHandler, this);
}
inline int INIReader::ParseError() const
{
return _error;
}
inline INIReader::~INIReader() { }
inline const std::set<std::string>& INIReader::Sections() const
{
return _sections;
}
inline std::string INIReader::Get(std::string section, std::string name, std::string default_value) const
{
std::string key = MakeKey(section, name);
return _values.count(key) ? _values.at(key) : default_value;
}
inline std::string INIReader::Get(std::string section, std::string name) const
{
std::string key = MakeKey(section, name);
if(_values.count(key)) return _values.at(key);
else
{
printf("[ERROR] Does not find the section %s with name %s. \n", section.c_str(), name.c_str());
exit(-1);
}
}
inline long INIReader::GetInteger(std::string section, std::string name, long default_value) const
{
std::string valstr = Get(section, name, "");
const char* value = valstr.c_str();
char* end;
// This parses "1234" (decimal) and also "0x4D2" (hex)
long n = strtol(value, &end, 0);
return end > value ? n : default_value;
}
inline long INIReader::GetInteger(std::string section, std::string name) const
{
std::string valstr = Get(section, name, "");
const char* value = valstr.c_str();
char* end;
// This parses "1234" (decimal) and also "0x4D2" (hex)
long n = strtol(value, &end, 0);
if(end <= value)
{
printf("[ERROR] Does not find the section %s with name %s. \n", section.c_str(), name.c_str());
exit(-1);
}
return n;
}
inline double INIReader::GetReal(std::string section, std::string name, double default_value) const
{
std::string valstr = Get(section, name, "");
const char* value = valstr.c_str();
char* end;
double n = strtod(value, &end);
return end > value ? n : default_value;
}
inline float INIReader::GetFloat(std::string section, std::string name, float default_value) const
{
std::string valstr = Get(section, name, "");
const char* value = valstr.c_str();
char* end;
float n = strtof(value, &end);
return end > value ? n : default_value;
}
inline float INIReader::GetFloat(std::string section, std::string name) const
{
std::string valstr = Get(section, name, "");
const char* value = valstr.c_str();
char* end;
float n = strtof(value, &end);
if(end <= value)
{
printf("[ERROR] Does not find the section %s with name %s. \n", section.c_str(), name.c_str());
exit(-1);
}
return n;
}
inline bool INIReader::GetBoolean(std::string section, std::string name, bool default_value) const
{
std::string valstr = Get(section, name, "");
// Convert to lower case to make string comparisons case-insensitive
std::transform(valstr.begin(), valstr.end(), valstr.begin(), ::tolower);
if (valstr == "true" || valstr == "yes" || valstr == "on" || valstr == "1")
return true;
else if (valstr == "false" || valstr == "no" || valstr == "off" || valstr == "0")
return false;
else
return default_value;
}
inline std::string INIReader::MakeKey(std::string section, std::string name)
{
std::string key = section + "=" + name;
// Convert to lower case to make section/name lookups case-insensitive
std::transform(key.begin(), key.end(), key.begin(), ::tolower);
return key;
}
inline int INIReader::ValueHandler(void* user, const char* section, const char* name,
const char* value)
{
INIReader* reader = (INIReader*)user;
std::string key = MakeKey(section, name);
if (reader->_values[key].size() > 0)
reader->_values[key] += "\n";
reader->_values[key] += value;
reader->_sections.insert(section);
return 1;
}
#endif // __INIREADER__
\ No newline at end of file
image: gitlab-master.nvidia.com/dl/dgx/pytorch:20.12-py3-devel
test:
script:
- pytest --junitxml=report.xml tests
artifacts:
when: always
reports:
junit: report.xml
\ No newline at end of file
The following applies to all files unless otherwise noted:
# Copyright (c) 2020, NVIDIA CORPORATION. All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions
# are met:
# * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# * Redistributions in binary form must reproduce the above copyright
# notice, this list of conditions and the following disclaimer in the
# documentation and/or other materials provided with the distribution.
# * Neither the name of NVIDIA CORPORATION nor the names of its
# contributors may be used to endorse or promote products derived
# from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
# OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
--
This repository also contains code from Hugging Face Inc., Google Research,
Facebook (from their Fairseq project), and Philip Popien. Files from these
organizations have notices at the top of each file. Below are licenses
used in those files, as indicated.
------------- LICENSE FOR huggingface and Google Research code --------------
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work
(an example is provided in the Appendix below).
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding those notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf
of any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
END OF TERMS AND CONDITIONS
APPENDIX: How to apply the Apache License to your work.
To apply the Apache License to your work, attach the following
boilerplate notice, with the fields enclosed by brackets "[]"
replaced with your own identifying information. (Don't include
the brackets!) The text should be enclosed in the appropriate
comment syntax for the file format. We also recommend that a
file or class name and description of purpose be included on the
same "printed page" as the copyright notice for easier
identification within third-party archives.
Copyright [yyyy] [name of copyright owner]
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.
------------- LICENSE FOR Facebook Fairseq code --------------
MIT License
Copyright (c) Facebook, Inc. and its affiliates.
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
Megatron ([1](https://arxiv.org/pdf/1909.08053.pdf) and [2](https://arxiv.org/pdf/2104.04473.pdf)) is a large, powerful transformer developed by the Applied Deep Learning Research team at NVIDIA. This repository is for ongoing research on training large transformer language models at scale. We developed efficient, model-parallel (tensor and pipeline), and multi-node pre-training of transformer based models such as [GPT](https://arxiv.org/abs/2005.14165), [BERT](https://arxiv.org/pdf/1810.04805.pdf), and [T5](https://arxiv.org/abs/1910.10683) using mixed precision.
Below are some of the projects where we have directly used Megatron:
* [BERT and GPT Studies Using Megatron](https://arxiv.org/pdf/1909.08053.pdf)
* [BioMegatron: Larger Biomedical Domain Language Model](https://www.aclweb.org/anthology/2020.emnlp-main.379.pdf)
* [End-to-End Training of Neural Retrievers for Open-Domain Question Answering](https://arxiv.org/abs/2101.00408)
* [Large Scale Multi-Actor Generative Dialog Modeling](https://www.aclweb.org/anthology/2020.acl-main.8.pdf)
* [Local Knowledge Powered Conversational Agents](https://arxiv.org/abs/2010.10150)
* [MEGATRON-CNTRL: Controllable Story Generation with External Knowledge Using Large-Scale Language Models](https://www.aclweb.org/anthology/2020.emnlp-main.226.pdf)
* [RACE Reading Comprehension Dataset Leaderboard](http://www.qizhexie.com/data/RACE_leaderboard.html)
* [Scaling Language Model Training to a Trillion Parameters Using Megatron](https://arxiv.org/pdf/2104.04473.pdf)
* [Training Question Answering Models From Synthetic Data](https://www.aclweb.org/anthology/2020.emnlp-main.468.pdf)
Megatron is also used in [NeMo Megatron](https://developer.nvidia.com/nvidia-nemo#nemo-megatron), a framework to help enterprises overcome the challenges of building and training sophisticated natural language processing models with billions and trillions of parameters.
Our codebase is capable of efficiently training very large (hundreds of billions of parameters) language models with both model and data parallelism. To demonstrate how the code scales with multiple GPUs and model sizes, we consider GPT models from 1 billion all the way to 1 trillion parameters. All models use a vocabulary size of 51,200 and a sequence length of 2048. We vary hidden size, number of attention heads, and number of layers to arrive at a specifc model size. As the model size increases, we also modestly increase the batch size. We leverage [NVIDIA's Selene supercomputer](https://www.top500.org/system/179842/) to perform scaling studies and use up to 3072 [A100](https://www.nvidia.com/en-us/data-center/a100/) GPUs for the largest model. Each cluster node has 8 NVIDIA 80GB A100 GPUs. The table below shows the model configurations along with the achieved FLOPs (both per GPU and aggregate over all GPUs). Note that these results are from benchmark runs and these models were not trained to convergence; however, the FLOPs are measured for end-to-end training, i.e., includes all operations including data loading, optimization, and even logging.
Additionally, the model parallel size column reports a combined tensor and pipeline parallelism degrees. For numbers larger than 8, typically tensor parallel of size 8 was used. So, for example, the 145B model reports the total model parallel size of 64, which means that this setup used TP=8 and PP=8.
![Cases](images/cases_april2021.png)
All the cases from 1 billion to 1 trillion parameters achieve more than 43% half precision utilization, which is high for an end-to-end application. We observe that initially the utilization remains constant but as hidden size increases for larger models, utilization starts increasing and reaches 52% for the largest model. We also note that achieved aggregate petaFLOPs across all GPUs increases almost linearly with number of GPUs, demonstrating good weak scaling.
# Contents
* [Contents](#contents)
* [Setup](#setup)
* [Downloading Checkpoints](#downloading-checkpoints)
* [Usage](#usage)
* [Training](#training)
* [Data Preprocessing](#data-preprocessing)
* [BERT Pretraining](#bert-pretraining)
* [GPT Pretraining](#gpt-pretraining)
* [T5 Pretraining](#t5-pretraining)
* [Distributed Pretraining](#distributed-pretraining)
* [GPT-3 Example](#gpt-3-example)
* [Evaluation and Tasks](#evaluation-and-tasks)
* [GPT Text Generation](#gpt-text-generation)
* [GPT Evaluation](#gpt-evaluation)
* [WikiText Perplexity Evaluation](#wikitext-perplexity-evaluation)
* [LAMBADA Cloze Accuracy](#lambada-cloze-accuracy)
* [BERT Task Evaluation](#bert-task-evaluation)
* [RACE Evaluation](#race-evaluation)
* [MNLI Evaluation](#mnli-evaluation)
* [Datasets](#datasets)
* [Collecting Wikipedia Training Data](#collecting-wikipedia-training-data)
* [Collecting GPT Webtext Data](#collecting-gpt-webtext-data)
# Setup
We have tested Megatron with [NGC's PyTorch container](https://ngc.nvidia.com/catalog/containers/nvidia:pytorch) version 20.12, which uses python 3.8, pytorch 1.8, cuda 11.1, and nccl 2.8.3.
To use this repository, please install the latest supported versions of PyTorch with GPU support (python 3.8, pytorch 1.8, cuda 11.1, and nccl 2.8.3 and above) and NVIDIA [APEX](https://github.com/NVIDIA/apex#quick-start). We strongly recommend using one of [NGC's recent PyTorch containers](https://ngc.nvidia.com/catalog/containers/nvidia:pytorch) (the latest compatible version at time of publication can be pulled with `docker pull nvcr.io/nvidia/pytorch:20.12-py3`). Data preprocessing requires [NLTK](https://www.nltk.org/install.html), though this is not required for training, evaluation, or downstream tasks.
## Downloading Checkpoints
We have provided pretrained [BERT-345M](https://ngc.nvidia.com/catalog/models/nvidia:megatron_bert_345m) and [GPT-345M](https://ngc.nvidia.com/catalog/models/nvidia:megatron_lm_345m) checkpoints for use to evaluate or finetuning downstream tasks. To access these checkpoints, first [sign up](https://ngc.nvidia.com/signup) for and [setup](https://ngc.nvidia.com/setup/installers/cli) the NVIDIA GPU Cloud (NGC) Registry CLI. Further documentation for downloading models can be found in the [NGC documentation](https://docs.nvidia.com/dgx/ngc-registry-cli-user-guide/index.html#topic_6_4_1).
Alternatively, you can directly download the checkpoints using:
<pre>
BERT-345M-uncased: wget --content-disposition https://api.ngc.nvidia.com/v2/models/nvidia/megatron_bert_345m/versions/v0.1_uncased/zip -O megatron_bert_345m_v0.1_uncased.zip
BERT-345M-cased: wget --content-disposition https://api.ngc.nvidia.com/v2/models/nvidia/megatron_bert_345m/versions/v0.1_cased/zip -O megatron_bert_345m_v0.1_cased.zip
GPT-345M: wget --content-disposition https://api.ngc.nvidia.com/v2/models/nvidia/megatron_lm_345m/versions/v0.0/zip -O megatron_lm_345m_v0.0.zip
</pre>
The models require vocabulary files to run. The BERT WordPiece vocab file can be extracted from Google's pretrained BERT models: [uncased](https://s3.amazonaws.com/models.huggingface.co/bert/bert-large-uncased-vocab.txt), [cased](https://s3.amazonaws.com/models.huggingface.co/bert/bert-large-cased-vocab.txt). The GPT [vocab file](https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-vocab.json) and [merge table](https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-merges.txt) can be downloaded directly.
# Usage
After installation, there are several possible workflows. The most comprehensive is:
1. Data preprocessing
2. Pretraining
3. Finetuning (Optional for zero-shot tasks)
4. Downstream task evaluation or text generation
However, steps 1 and 2 can be replaced by using one of the pretrained models mentioned above.
We've provided several scripts for pretraining both BERT and GPT in [`examples`](./examples) directory, as well as scripts for both zero-shot and fine-tuned downstream tasks including MNLI, RACE, WikiText103, and LAMBADA evaluation. There is also a script for GPT interactive text generation.
# Training
## Data Preprocessing
The training data requires preprocessing. First, place your training data in a loose json format, with one json containing a text sample per line. For example:
<pre>
{"src": "www.nvidia.com", "text": "The quick brown fox", "type": "Eng", "id": "0", "title": "First Part"}
{"src": "The Internet", "text": "jumps over the lazy dog", "type": "Eng", "id": "42", "title": "Second Part"}
</pre>
The name of the `text` field of the json can be changed by using the `--json-key` flag in [`preprocess_data.py`](./tools/preprocess_data.py) The other metadata are optional and are not used in training.
The loose json is then processed into a binary format for training. To convert the json into mmap, cached index file, or the lazy loader format use `preprocess_data.py`. Set the `--dataset-impl` flag to `mmap`, `cached`, or `lazy`, respectively (default is `mmap`). An example script to prepare data for BERT training is:
<pre>
python tools/preprocess_data.py \
--input my-corpus.json \
--output-prefix my-bert \
--vocab bert-vocab.txt \
--dataset-impl mmap \
--tokenizer-type BertWordPieceLowerCase \
--split-sentences
</pre>
The output will be two files named, in this case, `my-bert_text_sentence.bin` and `my-bert_text_sentence.idx`. The `--data-path` specified in later BERT training is the full path and new filename, but without the file extension.
For T5 use the same preprocessing as BERT, perhaps renaming it to:
<pre>
--output-prefix my-t5 \
</pre>
Some minor modifications are required for GPT data preprocessing, namely, the addition of a merge table, an end-of-document token, removal of sentence splitting, and a change to the tokenizer type:
<pre>
python tools/preprocess_data.py \
--input my-corpus.json \
--output-prefix my-gpt2 \
--vocab gpt2-vocab.json \
--dataset-impl mmap \
--tokenizer-type GPT2BPETokenizer \
--merge-file gpt2-merges.txt \
--append-eod
</pre>
Here the output files are named `my-gpt2_text_document.bin` and `my-gpt2_text_document.idx`. As before, in GPT training, use the longer name without the extension as `--data-path`.
Further command line arguments are described in the source file [`preprocess_data.py`](./tools/preprocess_data.py).
## BERT Pretraining
The `examples/pretrain_bert.sh` script runs single GPU 345M parameter BERT pretraining. Debugging is the primary use for single GPU training, as the code base and command line arguments are optimized for highly distributed training. Most of the arguments are fairly self-explanatory. By default, the learning rate decays linearly over the training iterations starting at `--lr` to a minimum set by `--min-lr` over `--lr-decay-iters` iterations. The fraction of training iterations used for warmup is set by `--lr-warmup-fraction`. While this is single GPU training, the batch size specified by `--micro-batch-size` is a single forward-backward path batch-size and the code will perform gradient accumulation steps until it reaches `global-batch-size` which is the batch size per iteration. The data is partitioned into a 949:50:1 ratio for training/validation/test sets (default is 969:30:1). This partitioning happens on the fly, but is consistent across runs with the same random seed (1234 by default, or specified manually with `--seed`). We use `train-iters` as the training iterations requested. Alternatively, one can provide `--train-samples` which is total number of samples to train on. If this option is present, then instead of providing `--lr-decay-iters`, one will need to provide `--lr-decay-samples`.
The logging, checkpoint-saving, and evaluation intervals are specified. Checkpointing the activations facilitates the training of larger models and/or batches. Note that the `--data-path` now includes the additional `_text_sentence` suffix added in preprocessing, but does not include the file extensions.
<pre>
CHECKPOINT_PATH=checkpoints/bert_345m
VOCAB_FILE=bert-vocab.txt
DATA_PATH=my-bert_text_sentence
BERT_ARGS="--num-layers 24 \
--hidden-size 1024 \
--num-attention-heads 16 \
--seq-length 512 \
--max-position-embeddings 512 \
--lr 0.0001 \
--lr-decay-iters 990000 \
--train-iters 2000000 \
--min-lr 0.00001 \
--lr-warmup-fraction 0.01 \
--micro-batch-size 4 \
--global-batch-size 8 \
--vocab-file $VOCAB_FILE \
--split 949,50,1 \
--fp16"
OUTPUT_ARGS="--log-interval 10 \
--save-interval 500 \
--eval-interval 100 \
--eval-iters 10 \
--activations-checkpoint-method uniform"
python pretrain_bert.py \
$BERT_ARGS \
$OUTPUT_ARGS \
--save $CHECKPOINT_PATH \
--load $CHECKPOINT_PATH \
--data-path $DATA_PATH
</pre>
Further command line arguments are described in the source file [`arguments.py`](./megatron/arguments.py).
## GPT Pretraining
The `examples/pretrain_gpt.sh` script runs single GPU 345M parameter GPT pretraining. As mentioned above, single GPU training is primarily intended for debugging purposes, as the code is optimized for distributed training.
It follows largely the same format as the previous BERT script with a few notable differences: the tokenization scheme used is BPE (which requires a merge table and a `json` vocabulary file) instead of WordPiece, the model architecture allows for longer sequences (note that the max position embedding must be greater than or equal to the maximum sequence length), and the `--lr-decay-style` has been set to cosine decay. Note that the `--data-path` now includes the additional `_text_document` suffix added in preprocessing, but does not include the file extensions.
<pre>
CHECKPOINT_PATH=checkpoints/gpt2_345m
VOCAB_FILE=gpt2-vocab.json
MERGE_FILE=gpt2-merges.txt
DATA_PATH=my-gpt2_text_document
GPT_ARGS="--num-layers 24 \
--hidden-size 1024 \
--num-attention-heads 16 \
--seq-length 1024 \
--max-position-embeddings 1024 \
--micro-batch-size 4 \
--global-batch-size 8 \
--lr 0.00015 \
--train-iters 500000 \
--lr-decay-iters 320000 \
--lr-decay-style cosine \
--vocab-file $VOCAB_FILE \
--merge-file $MERGE_FILE \
--lr-warmup-fraction .01 \
--fp16"
OUTPUT_ARGS=&#60;same as those in <a href="#bert-pretraining">BERT pretraining</a> above&#62;
python pretrain_gpt.py \
$GPT_ARGS \
$OUTPUT_ARGS \
--save $CHECKPOINT_PATH \
--load $CHECKPOINT_PATH \
--data-path $DATA_PATH \
</pre>
Further command line arguments are described in the source file [`arguments.py`](./megatron/arguments.py).
## T5 Pretraining
Very similar to BERT and GPT, the `examples/pretrain_t5.sh` script runs single GPU "base" (~220M parameter) T5 pretraining. The primary difference from BERT and GPT is the addition of the following arguments to accommodate the T5 architecture:
* `--kv-channels` sets the inner dimension of the "key" and "value" matrices of all attention mechanisms in the model. For BERT and GPT this defaults to the hidden size divided by the number of attention heads, but can be configured for T5.
* `--ffn-hidden-size` sets the hidden size in the feed-forward networks within a transformer layer. For BERT and GPT this defaults to 4 times the transformer hidden size, but can be configured for T5.
* `--encoder-seq-length` and `--decoder-seq-length` set the sequence length for the encoder and decoder separately.
All of the other arguments remain as they were for BERT and GPT pretraining.
<pre>
CHECKPOINT_PATH=checkpoints/t5_base
VOCAB_FILE=t5-vocab.txt
DATA_PATH=my-t5_text_sentence
T5_ARGS="--num-layers 24 \
--hidden-size 1024 \
--num-attention-heads 16 \
--kv-channels 64 \
--ffn-hidden-size 3072 \
--encoder-seq-length 512 \
--decoder-seq-length 128 \
--max-position-embeddings 512 \
--lr 0.0001 \
--lr-decay-iters 990000 \
--train-iters 2000000 \
--min-lr 0.00001 \
--lr-warmup-fraction 0.01 \
--micro-batch-size 16 \
--global-batch-size 2048 \
--vocab-file $VOCAB_FILE \
--vocab-extra-ids 100 \
--split 949,50,1 \
--fp16"
OUTPUT_ARGS=&#60;same as those in <a href="#bert-pretraining">BERT pretraining</a> above&#62;
python pretrain_t5.py \
$T5_ARGS \
$OUTPUT_ARGS \
--save $CHECKPOINT_PATH \
--load $CHECKPOINT_PATH \
--data-path $DATA_PATH
</pre>
## Distributed Pretraining
The `examples/pretrain_{bert,gpt,t5}_distributed.sh` scripts use the PyTorch distributed launcher for distributed training. As such, multi-node training can be achieved by properly setting environment variables and using `init_method='env://'` in the launcher. See the official PyTorch [documentation](https://pytorch.org/docs/stable/distributed.html#launch-utility) for further description of these [environment variables](https://pytorch.org/docs/stable/distributed.html#environment-variable-initialization). By default, multi-node training uses the [nccl](https://developer.nvidia.com/nccl) distributed backend. A simple set of additional arguments and the use of the PyTorch distributed module with the Python flag `-m torch.distributed.launch`, detailed below, are the only additional requirements to adopt distributed training.
We use two types of parallelism: data and model parallelism. We facilitate two distributed data parallel implementations: a simple one of our own that performs gradient all-reduce at the end of back propagation step, and Torch's distributed data parallel wrapper that overlaps gradient reduction with back propagation computation. To switch between these two options use `--DDP-impl local` or `--DDP-impl torch`, respectively. As expected, Torch distributed data parallelism is more efficient at larger model sizes. For example, for the 8.3 billion parameters model running on 512 GPUs, the scaling increases from 60% to 76% when Torch's distributed data parallel is used. However, the overlapping method requires more memory and for some configurations (e.g., 2.5 billion parameters using 2-way model parallel and 1.2 billion parameters with no model parallel) can make the overall training slower as a result. We empirically found that using a smaller model in those cases improves the training time.
Second, we developed a simple and efficient two-dimensional model-parallel approach. To use tensor model parallelism (splitting execution of a single transformer module over multiple GPUs), add the `--tensor-model-parallel-size` flag to specify the number of GPUs among which to split the model, along with the arguments passed to the distributed launcher as mentioned above. To use pipeline model parallelism (sharding the transformer modules into stages with an equal number of transformer modules on each stage, and then pipelining execution by breaking the batch into smaller microbatches), use the `--pipeline-model-parallel-size` flag to specify the number of stages to split the model into (e.g., splitting a model with 24 transformer layers across 4 stages would mean each stage gets 6 transformer layers each).
<!-- The number of microbatches in a per-pipeline minibatch is controlled by the `--num-microbatches-in-minibatch` argument. With `WORLD_SIZE` GPUs, `TENSOR_MP_SIZE` tensor-model-parallel size, `PIPELINE_MP_SIZE` pipeline-model-parallel-size, `WORLD_SIZE`/(`TENSOR_MP_SIZE` * `PIPELINE_MP_SIZE`) GPUs will be used for data parallelism. The default values for `--tensor-model-parallel-size` and `--pipeline-model-parallel-size` is 1, which will not implement either form of model parallelism. -->
We have examples of how to use these two different forms of model parallelism the example scripts ending in `distributed_with_mp.sh`:
Other than these minor changes, the distributed training is identical to the training on a single GPU.
Distributed training:
<pre>
WORLD_SIZE=8
TENSOR_MP_SIZE=2
PIPELINE_MP_SIZE=2
DISTRIBUTED_ARGS="--nproc_per_node $WORLD_SIZE \
--nnodes 1 \
--node_rank 0 \
--master_addr localhost \
--master_port 6000"
CHECKPOINT_PATH=&#60;same as above&#62;
VOCAB_FILE=&#60;same as above&#62;
DATA_PATH=&#60;same as above&#62;
MODEL_ARGS=&#60;same as above&#62;
OUTPUT_ARGS=&#60;same as above&#62;
python -m torch.distributed.launch $DISTRIBUTED_ARGS ./pretrain_<model>.py \
$MODEL_ARGS \
$OUTPUT_ARGS \
--save $CHECKPOINT_PATH \
--load $CHECKPOINT_PATH \
--data-path $DATA_PATH \
--tensor-model-parallel-size $TENSOR_MP_SIZE \
--pipeline-model-parallel-size $PIPELINE_MP_SIZE \
--DDP-impl torch
</pre>
The interleaved pipelining schedule (more details in Section 2.2.2 of [our paper](https://arxiv.org/pdf/2104.04473.pdf)) can be enabled using the `--num-layers-per-virtual-pipeline-stage` argument, which controls the number of transformer layers in a virtual stage (by default with the non-interleaved schedule, each GPU will execute a single virtual stage with `NUM_LAYERS / PIPELINE_MP_SIZE` transformer layers). The total number of layers in the transformer model should be divisible by this argument value. Additionally, the number of microbatches in the pipeline (computed as `GLOBAL_BATCH_SIZE / (DATA_PARALLEL_SIZE * MICRO_BATCH_SIZE)`) should be divisible by the `PIPELINE_MP_SIZE` when using this schedule (this condition is checked in an assertion in the code). The interleaved schedule is not supported for pipelines with 2 stages (`PIPELINE_MP_SIZE=2`).
## Activation Checkpointing and Recomputation
To reduce GPU memory usage so deploy a large model to a training system, we support activation checkpointing and recomputation. We use a Transformer layer as the unit of checkpointing because the activation size bloats in the middle of a Transformer layer so checkpointing the input of a Transformer layer is storage-efficient. We support two activation checkpointing methods: `uniform` and `block`.
Uniform method uniformly divides the Transformer layers into groups of layers and stores the input activations of each group in the memory. The baseline group size is 1 and, in this case, the input activation of each Transformer layer is checkpointed. When the GPU memory is insufficient, increasing the number of layers per group reduces the memory usage thus enables running a bigger model. For example, when using the number of layers per group of 4, the input activation of each group of 4 Transformer layers is checkpointed.
Block method checkpoints the input activations of a set number of individual Transformer layers per pipeline stage and do the rest of layers without any checkpointing. This method can be used to skip checkpointing some Transformer layers until the GPU memory is fully used, which is applicable only when there is unused GPU memory. Checkpointing fewer transformer layers avoids unnecessary activation recomputation in the backprop thus improves training performance. For example, when we specify 5 layers to checkpoint of 8 layers per pipeline stage, the input activations of only the first 5 Transformer layers are checkpointed and activation recomputation for the rest 3 layers is not needed in the backprop.
## GPT-3 Example
In `examples/pretrain_gpt3_175B.sh` we have provided an example of how to configure Megatron to run [GPT-3](https://arxiv.org/abs/2005.14165) with 175 billion parameters on 1024 GPUs. The script is designed for [slurm](https://slurm.schedmd.com/documentation.html) with [pyxis](https://github.com/NVIDIA/pyxis) plugin but can be easily adopted to any other scheduler. It uses 8-way and 16-way tensor and pipeline parallelism, respectively. With options `global-batch-size 1536` and `rampup-batch-size 16 16 5859375`, the training will start with global batch size 16 and linearly increase the global batch size to 1536 over 5,859,375 samples with incrmeental steps 16. The training dataset can be either a single set or a multiple datasets combined with a set of weights.
With full global batch size of 1536 on 1024 A100 GPUs, each iteration takes around 32 seconds resulting in 138 teraFLOPs per GPU which is 44% of the theoretical peak FLOPs.
<!--
## REALM Pipeline
We are working on implementing the [REALM](https://arxiv.org/pdf/2002.08909.pdf) system. The following sections (will) reflect the three stages of training it. For now it's just the ICT code.
Loosely, they are pretraining the retriever modules, then jointly training the language model and the retriever, and then finetuning a question answering head on the language model with fixed retriever.
### Inverse Cloze Task (ICT) Pretraining
1. Have a corpus in loose JSON format with the intention of creating a collection of fixed-size blocks of text as the fundamental units of data. For a corpus like Wikipedia, this will mean multiple sentences per block but also multiple blocks per document.
Run `tools/preprocess_data.py` to construct one or more indexed datasets with the `--split-sentences` argument to make sentences the basic unit. For the original REALM system, we construct two datasets, one with the title of every document, and another with the body.
Refer to the following script
<pre>
python preprocess_data.py \
--input /path/to/corpus.json \
--json-keys text title \
--split-sentences \
--tokenizer-type BertWordPieceLowerCase \
--vocab-file /path/to/vocab.txt \
--output-prefix corpus_indexed \
--workers 5 # works well for 10 CPU cores. Scale up accordingly.
</pre>
2. Use a custom samples mapping function in place of `megatron/data/realm_dataset_utils.get_block_samples_mapping` if required. To do this, you will need to implement a new function in C++ inside of `megatron/data/helpers.cpp`. The samples mapping data structure is used to select the data that will constitute every training sample in advance of the training loop.
The samples mapping is responsible for holding all of the required metadata needed to construct the sample from one or more indexed datasets. In REALM, the samples mapping contains the start and end sentence indices, as well as the document index (to find the correct title for a body) and a unique ID for every block.
3. Pretrain a BERT language model using `pretrain_bert.py`, with the sequence length equal to the block size in token ids. This model should be trained on the same indexed dataset that is used to supply the blocks for the information retrieval task.
In REALM, this is an uncased bert base model trained with the standard hyperparameters.
4. Use `pretrain_ict.py` to train an `ICTBertModel` which uses two BERT-based encoders to encode queries and blocks to perform retrieval with.
The script below trains the ICT model from REALM. It refrences a pretrained BERT model (step 3) in the `--bert-load` argument. The batch size used in the paper is 4096, so this would need to be run with data parallel world size 32.
<pre>
python pretrain_ict.py \
--num-layers 12 \
--num-attention-heads 12 \
--hidden-size 768 \
--batch-size 128 \
--seq-length 256 \
--max-position-embeddings 256 \
--ict-head-size 128 \
--train-iters 100000 \
--activations-checkpoint-method uniform \
--bert-load /path/to/pretrained_bert \
--load checkpoints \
--save checkpoints \
--data-path /path/to/indexed_dataset \
--titles-data-path /path/to/titles_indexed_dataset \
--vocab-file /path/to/vocab.txt \
--lr 0.0001 \
--num-workers 2 \
--lr-decay-style linear \
--weight-decay 1e-2 \
--clip-grad 1.0 \
--lr-warmup-fraction .01 \
--save-interval 3000 \
--query-in-block-prob 0.1 \
--fp16
</pre>
### Building an Index of Block Embeddings
After having trained an ICT model, you can now embed an entire dataset of blocks by creating a `BlockData` structure. After that has been saved, you can load it
and wrap it with a `FaissMIPSIndex` to do fast similarity search which is key in the learned information retrieval pipeline. The initial index can be built with the following script, meant to be run in an interactive session. It can leverage multiple GPUs on multiple nodes to index large datasets much more quickly.
<pre>
python tools/create_doc_index.py \
--num-layers 12 \
--hidden-size 768 \
--ict-head-size 128 \
--num-attention-heads 12 \
--batch-size 128 \
--activations-checkpoint-method uniform \
--seq-length 256 \
--max-position-embeddings 256 \
--ict-load /path/to/pretrained_ict \
--data-path /path/to/indexed_dataset \
--titles-data-path /path/to/titles_indexed_dataset \
--block-data-path embedded_blocks.pkl \
--indexer-log-interval 1000 \
--indexer-batch-size 128 \
--vocab-file /path/to/vocab.txt \
--num-workers 2 \
--fp16
</pre>
-->
# Evaluation and Tasks
We provide several command line arguments, detailed in the scripts listed below, to handle various zero-shot and fine-tuned downstream tasks. However, you can also finetune your model from a pretrained checkpoint on other corpora as desired. To do so, simply add the `--finetune` flag and adjust the input files and training parameters within the original training script. The iteration count will be reset to zero, and the optimizer and internal state will be reinitialized. If the fine-tuning is interrupted for any reason, be sure to remove the `--finetune` flag before continuing, otherwise the training will start again from the beginning.
Because evaluation requires substantially less memory than training, it may be advantageous to merge a model trained in parallel for use on a single GPU in downstream tasks. The following script accomplishes this. Currently only tensor model parallelism is supported on input and pipeline model parallelism on the output. This example reads in a model with 2-way tensor model parallelism and writes out a model with 2-way pipeline model parallelism.
<pre>
TENSOR_MODEL_PARALLEL_SIZE=2
TARGET_PIPELINE_MODEL_PARALLEL_SIZE=2
VOCAB_FILE=bert-vocab.txt
CHECKPOINT_PATH=checkpoints/bert_345m
WORLD_SIZE=$TENSOR_MODEL_PARALLEL_SIZE python tools/merge_mp_partitions.py \
--model-type BERT \
--tensor-model-parallel-size $TENSOR_MODEL_PARALLEL_SIZE \
--pipeline-model-parallel-size 1 \
--target-pipeline-model-parallel-size $TARGET_PIPELINE_MODEL_PARALLEL_SIZE \
--tokenizer-type BertWordPieceLowerCase \
--vocab-file $VOCAB_FILE \
--num-layers 24 \
--hidden-size 1024 \
--num-attention-heads 16 \
--seq-length 512 \
--max-position-embeddings 512 \
--load $CHECKPOINT_PATH
--save $CHECKPOINT_PATH/merged
</pre>
Several downstream tasks are described for both GPT and BERT models below. They can be run in distributed and model parallel modes with the same changes used in the training scripts.
## GPT Text Generation
We have included a simple REST server to use for text generation in `tools/run_text_generation_server.py`. You run it much like you would start a pretraining job, specifying an appropriate pretrained checkpoint. There are also few optional parameters: `temperature`, `top-k`and `top-p`. See `--help` or the source file for more information. See [examples/run_text_generation_server_345M.sh](examples/run_text_generation_server_345M.sh) for an example of how to run the server.
Once the server is running you can use `tools/text_generation_cli.py` to query it, it takes one argument which is the host the server is running on.
<pre>
tools/text_generation_cli.py localhost
</pre>
You can also use CURL or any other tools to query the server directly:
<pre>
curl 'http://localhost:5000/api' -X 'PUT' -H 'Content-Type: application/json; charset=UTF-8' -d '{"prompts":["Hello world"], "tokens_to_generate":1}'
</pre>
See [megatron/text_generation_server.py](megatron/text_generation_server.py) for more API options.
## GPT Evaluation
We include example scripts for GPT evaluation on WikiText perplexity evaluation and LAMBADA Cloze accuracy.
### WikiText Perplexity Evaluation
For even comparison with prior works, we evaluate perplexity on the word-level [WikiText-103 test dataset](https://s3.amazonaws.com/research.metamind.io/wikitext/wikitext-103-v1.zip), and appropriately compute perplexity given the change in tokens when using our subword tokenizer.
We use the following command to run WikiText-103 evaluation on a 345M parameter model.
<pre>
TASK="WIKITEXT103"
VALID_DATA=&#60;wikitext path&#62;.txt
VOCAB_FILE=gpt2-vocab.json
MERGE_FILE=gpt2-merges.txt
CHECKPOINT_PATH=checkpoints/gpt2_345m
COMMON_TASK_ARGS="--num-layers 24 \
--hidden-size 1024 \
--num-attention-heads 16 \
--seq-length 1024 \
--max-position-embeddings 1024 \
--fp16 \
--vocab-file $VOCAB_FILE"
python tasks/main.py \
--task $TASK \
$COMMON_TASK_ARGS \
--valid-data $VALID_DATA \
--tokenizer-type GPT2BPETokenizer \
--merge-file $MERGE_FILE \
--load $CHECKPOINT_PATH \
--micro-batch-size 8 \
--activations-checkpoint-method uniform \
--log-interval 10 \
--no-load-optim \
--no-load-rng
</pre>
### LAMBADA Cloze Accuracy
To compute LAMBADA cloze accuracy (the accuracy of predicting the last token given the preceding tokens) we utilize a detokenized, processed version of the [LAMBADA dataset](https://github.com/cybertronai/bflm/blob/master/lambada_test.jsonl).
We use the following command to run LAMBADA evaluation on a 345M parameter model. Note that the `--strict-lambada` flag should be used to require whole word matching. Make that `lambada` is part of the file path.
<pre>
TASK="LAMBADA"
VALID_DATA=&#60;lambada path&#62;.json
VOCAB_FILE=gpt2-vocab.json
MERGE_FILE=gpt2-merges.txt
CHECKPOINT_PATH=checkpoints/gpt2_345m
COMMON_TASK_ARGS=&#60;same as those in <a href="#wikitext-perplexity-evaluation">WikiText Perplexity Evaluation</a> above&#62;
python tasks/main.py \
--task $TASK \
$COMMON_TASK_ARGS \
--valid-data $VALID_DATA \
--tokenizer-type GPT2BPETokenizer \
--strict-lambada \
--merge-file $MERGE_FILE \
--load $CHECKPOINT_PATH \
--micro-batch-size 8 \
--activations-checkpoint-method uniform \
--log-interval 10 \
--no-load-optim \
--no-load-rng
</pre>
Further command line arguments are described in the source file [`main.py`](./tasks/main.py)
## BERT Task Evaluation
### RACE Evaluation
The following script finetunes the BERT model for evaluation on the [RACE dataset](http://www.cs.cmu.edu/~glai1/data/race/). The `TRAIN_DATA` and `VALID_DATA` directory contain the RACE dataset as separate `.txt` files. Note that for RACE, the batch size is the number of RACE query's to evaluate. Since each RACE query has four samples, the effective batch size passed through the model will be four times the batch size specified on the command line.
<pre>
TRAIN_DATA="data/RACE/train/middle"
VALID_DATA="data/RACE/dev/middle \
data/RACE/dev/high"
VOCAB_FILE=bert-vocab.txt
PRETRAINED_CHECKPOINT=checkpoints/bert_345m
CHECKPOINT_PATH=checkpoints/bert_345m_race
COMMON_TASK_ARGS="--num-layers 24 \
--hidden-size 1024 \
--num-attention-heads 16 \
--seq-length 512 \
--max-position-embeddings 512 \
--fp16 \
--vocab-file $VOCAB_FILE"
COMMON_TASK_ARGS_EXT="--train-data $TRAIN_DATA \
--valid-data $VALID_DATA \
--pretrained-checkpoint $PRETRAINED_CHECKPOINT \
--activations-checkpoint-method uniform \
--save-interval 10000 \
--save $CHECKPOINT_PATH \
--log-interval 100 \
--eval-interval 1000 \
--eval-iters 10 \
--weight-decay 1.0e-1"
python tasks/main.py \
--task RACE \
$COMMON_TASK_ARGS \
$COMMON_TASK_ARGS_EXT \
--tokenizer-type BertWordPieceLowerCase \
--epochs 3 \
--micro-batch-size 4 \
--lr 1.0e-5 \
--lr-warmup-fraction 0.06
</pre>
### MNLI Evaluation
The following script finetunes the BERT model for evaluation with the [MultiNLI sentence pair corpus](https://www.nyu.edu/projects/bowman/multinli/). Because the matching tasks are quite similar, the script can be quickly tweaked to work with the [Quora Question Pairs](https://www.kaggle.com/quora/question-pairs-dataset) (QQP) dataset as well.
<pre>
TRAIN_DATA="data/glue_data/MNLI/train.tsv"
VALID_DATA="data/glue_data/MNLI/dev_matched.tsv \
data/glue_data/MNLI/dev_mismatched.tsv"
PRETRAINED_CHECKPOINT=checkpoints/bert_345m
VOCAB_FILE=bert-vocab.txt
CHECKPOINT_PATH=checkpoints/bert_345m_mnli
COMMON_TASK_ARGS=&#60;same as those in <a href="#race-evaluation">RACE Evaluation</a> above&#62;
COMMON_TASK_ARGS_EXT=&#60;same as those in <a href="#race-evaluation">RACE Evaluation</a> above&#62;
python tasks/main.py \
--task MNLI \
$COMMON_TASK_ARGS \
$COMMON_TASK_ARGS_EXT \
--tokenizer-type BertWordPieceLowerCase \
--epochs 5 \
--micro-batch-size 8 \
--lr 5.0e-5 \
--lr-warmup-fraction 0.065
</pre>
# Datasets
We do not host any datasets for GPT or BERT training, however, we detail their collection so that our results may be reproduced.
## Collecting Wikipedia Training Data
We recommend following the Wikipedia data extraction process specified by Google research: "the recommended pre-processing is to download [the latest dump](https://dumps.wikimedia.org/enwiki/latest/enwiki-latest-pages-articles.xml.bz2), extract the text with [WikiExtractor.py](https://github.com/attardi/wikiextractor), and then apply any necessary cleanup to convert it into plain text."
We recommend using the `--json` argument when using WikiExtractor, which will dump the Wikipedia data into loose json format (one json per line), making it more manageable on the file system and also readily consumable by our codebase. We recommend further preprocessing this json dataset by nltk punctuation standardization. For BERT training, use the `--split-sentences` flag to `preprocess_data.py` as described [above](#data-preprocessing) to include sentence breaks in the produced index. If you'd like to use Wikipedia data for GPT training you should still clean it with nltk/spacy/ftfy, but do not use the `--split-sentences` flag.
## Collecting GPT Webtext Data
We utilize the publicly available [OpenWebText](https://github.com/eukaryote31/openwebtext) library from [jcpeterson](https://github.com/jcpeterson/openwebtext) and [eukaryote31's](https://github.com/eukaryote31/openwebtext) work to download urls. We then filtered, cleaned, and deduplicated all downloaded content according to the procedure described in our [openwebtext](./tools/openwebtext) directory. For reddit URLs corresponding to content up to October 2018 we arrived at approximately 37GB of content.
#!/bin/bash
# Evaluate natural question test data given Wikipedia embeddings and pretrained
# ICT model or a finetuned model for Natural Question task
# Datasets can be downloaded from the following link:
# https://github.com/facebookresearch/DPR/blob/master/data/download_data.py
EVIDENCE_DATA_DIR=<Specify path of Wikipedia dataset>
EMBEDDING_PATH=<Specify path of the embeddings>
CHECKPOINT_PATH=<Specify path of pretrained ICT model or finetuned model>
QA_FILE=<Path of the natural question dev or test dataset>
python tasks/main.py \
--task RETRIEVER-EVAL \
--tokenizer-type BertWordPieceLowerCase \
--num-layers 12 \
--hidden-size 768 \
--num-attention-heads 12 \
--tensor-model-parallel-size 1 \
--micro-batch-size 128 \
--activations-checkpoint-method uniform \
--seq-length 512 \
--max-position-embeddings 512 \
--load ${CHECKPOINT_PATH} \
--evidence-data-path ${EVIDENCE_DATA_DIR} \
--embedding-path ${EMBEDDING_PATH} \
--retriever-seq-length 256 \
--vocab-file bert-vocab.txt\
--qa-data-test ${QA_FILE} \
--faiss-use-gpu \
--retriever-report-topk-accuracies 1 5 20 100 \
--fp16 \
--indexer-log-interval 1000 \
--indexer-batch-size 128
#!/bin/bash
WORLD_SIZE=8
DISTRIBUTED_ARGS="--nproc_per_node $WORLD_SIZE \
--nnodes 1 \
--node_rank 0 \
--master_addr localhost \
--master_port 6000"
TASK="LAMBADA"
VALID_DATA=<lambada path>
VOCAB_FILE=gpt2-vocab.json
MERGE_FILE=gpt2-merges.txt
CHECKPOINT=checkpoints/gpt2_345m
python -m torch.distributed.launch $DISTRIBUTED_ARGS ./tasks/main.py \
--task $TASK \
--valid-data $VALID_DATA \
--tokenizer-type GPT2BPETokenizer \
--strict-lambada \
--vocab-file $VOCAB_FILE \
--merge-file $MERGE_FILE \
--load $CHECKPOINT \
--tensor-model-parallel-size 1 \
--num-layers 24 \
--hidden-size 1024 \
--num-attention-heads 16 \
--batch-size 8 \
--activations-checkpoint-method uniform \
--seq-length 1024 \
--max-position-embeddings 1024 \
--log-interval 10 \
--fp16 \
--no-load-optim \
--no-load-rng
#!/bin/bash
WORLD_SIZE=8
DISTRIBUTED_ARGS="--nproc_per_node $WORLD_SIZE \
--nnodes 1 \
--node_rank 0 \
--master_addr localhost \
--master_port 6000"
TRAIN_DATA="data/glue_data/MNLI/train.tsv"
VALID_DATA="data/glue_data/MNLI/dev_matched.tsv \
data/glue_data/MNLI/dev_mismatched.tsv"
PRETRAINED_CHECKPOINT=checkpoints/bert_345m
VOCAB_FILE=bert-vocab.txt
CHECKPOINT_PATH=checkpoints/bert_345m_mnli
python -m torch.distributed.launch $DISTRIBUTED_ARGS ./tasks/main.py \
--task MNLI \
--seed 1234 \
--train-data $TRAIN_DATA \
--valid-data $VALID_DATA \
--tokenizer-type BertWordPieceLowerCase \
--vocab-file $VOCAB_FILE \
--epochs 5 \
--pretrained-checkpoint $PRETRAINED_CHECKPOINT \
--tensor-model-parallel-size 1 \
--num-layers 24 \
--hidden-size 1024 \
--num-attention-heads 16 \
--micro-batch-size 8 \
--activations-checkpoint-method uniform \
--lr 5.0e-5 \
--lr-decay-style linear \
--lr-warmup-fraction 0.065 \
--seq-length 512 \
--max-position-embeddings 512 \
--save-interval 500000 \
--save $CHECKPOINT_PATH \
--log-interval 10 \
--eval-interval 100 \
--eval-iters 50 \
--weight-decay 1.0e-1 \
--fp16
#!/bin/bash
WORLD_SIZE=8
DISTRIBUTED_ARGS="--nproc_per_node $WORLD_SIZE \
--nnodes 1 \
--node_rank 0 \
--master_addr localhost \
--master_port 6000"
TRAIN_DATA="data/RACE/train/middle"
VALID_DATA="data/RACE/dev/middle \
data/RACE/dev/high"
VOCAB_FILE=bert-vocab.txt
PRETRAINED_CHECKPOINT=checkpoints/bert_345m
CHECKPOINT_PATH=checkpoints/bert_345m_race
python -m torch.distributed.launch $DISTRIBUTED_ARGS ./tasks/main.py \
--task RACE \
--seed 1234 \
--train-data $TRAIN_DATA \
--valid-data $VALID_DATA \
--tokenizer-type BertWordPieceLowerCase \
--vocab-file $VOCAB_FILE \
--epochs 3 \
--pretrained-checkpoint $PRETRAINED_CHECKPOINT \
--tensor-model-parallel-size 1 \
--num-layers 24 \
--hidden-size 1024 \
--num-attention-heads 16 \
--micro-batch-size 4 \
--activations-checkpoint-method uniform \
--lr 1.0e-5 \
--lr-decay-style linear \
--lr-warmup-fraction 0.06 \
--seq-length 512 \
--max-position-embeddings 512 \
--save-interval 100000 \
--save $CHECKPOINT_PATH \
--log-interval 10 \
--eval-interval 100 \
--eval-iters 50 \
--weight-decay 1.0e-1 \
--clip-grad 1.0 \
--hidden-dropout 0.1 \
--attention-dropout 0.1 \
--fp16
#!/bin/bash
# Finetune a BERT or pretrained ICT model using Google natural question data
# Datasets can be downloaded from the following link:
# https://github.com/facebookresearch/DPR/blob/master/data/download_data.py
WORLD_SIZE=8
DISTRIBUTED_ARGS="--nproc_per_node $WORLD_SIZE \
--nnodes 1 \
--node_rank 0 \
--master_addr localhost \
--master_port 6000"
CHECKPOINT_PATH=<Specify path for the finetuned retriever model>
# Load either of the below
BERT_LOAD_PATH=<Path of BERT pretrained model>
PRETRAINED_CHECKPOINT=<Path of Pretrained ICT model>
python -m torch.distributed.launch $DISTRIBUTED_ARGS ./tasks/main.py \
--task RET-FINETUNE-NQ \
--train-with-neg \
--train-hard-neg 1 \
--pretrained-checkpoint ${PRETRAINED_CHECKPOINT} \
--num-layers 12 \
--hidden-size 768 \
--num-attention-heads 12 \
--tensor-model-parallel-size 1 \
--tokenizer-type BertWordPieceLowerCase \
--train-data nq-train.json \
--valid-data nq-dev.json \
--save ${CHECKPOINT_PATH} \
--load ${CHECKPOINT_PATH} \
--vocab-file bert-vocab.txt \
--bert-load ${BERT_LOAD_PATH} \
--save-interval 5000 \
--log-interval 10 \
--eval-interval 20000 \
--eval-iters 100 \
--indexer-log-interval 1000 \
--faiss-use-gpu \
--DDP-impl torch \
--fp16 \
--retriever-report-topk-accuracies 1 5 10 20 100 \
--seq-length 512 \
--retriever-seq-length 256 \
--max-position-embeddings 512 \
--retriever-score-scaling \
--epochs 80 \
--micro-batch-size 8 \
--eval-micro-batch-size 16 \
--indexer-batch-size 128 \
--lr 2e-5 \
--lr-warmup-fraction 0.01 \
--weight-decay 1e-1
#!/bin/bash
TENSOR_MODEL_PARALLEL_SIZE=2
VOCAB_FILE=bert-vocab.txt
CHECKPOINT_PATH=checkpoints/bert_345m
WORLD_SIZE=$TENSOR_MODEL_PARALLEL_SIZE python tools/merge_mp_partitions.py \
--model-type BERT \
--tensor-model-parallel-size $TENSOR_MODEL_PARALLEL_SIZE \
--tokenizer-type BertWordPieceLowerCase \
--vocab-file $VOCAB_FILE \
--num-layers 24 \
--hidden-size 1024 \
--num-attention-heads 16 \
--seq-length 512 \
--max-position-embeddings 512 \
--load $CHECKPOINT_PATH
# Multi-Stage Prompting for Knowledgeable Dialogue Generation
This directory contains all the scripts of multi-stage prompting for knowledgeable dialogue generation that includes data preparation, and knowledge and response generations. More details are available on [`knowledgeable task directory`](../../tasks/msdp).
#!/bin/bash
# Data preparation for our framework: preprocessing the WoW and WoI datasets
# The datasets can be downloaded through the following links:
# WoW: https://parl.ai/projects/wizard_of_wikipedia/
# WoI: https://parl.ai/projects/sea/
DIR=`pwd`
# Before running the preprocessing, please download
# the wizard of wikipedia and wizard datasets
WOW_DATA_FOLDER=<PATH_OF_WIZARD_OF_WIKIPEDIA_DATA_FOLDER>
WOI_DATA_FOLDER=<PATH_OF_WIZARD_OF_INTERNET_DATA_FOLDER>
# We provide examples for processing the raw data from Wizard of Wikipedia
# Processing the train dataset (train.json)
python ${DIR}/tasks/msdp/preprocessing.py \
--func process_wow_dataset \
--raw_file ${WOW_DATA_FOLDER}/train.json \
--processed_file ${WOW_DATA_FOLDER}/train_processed.txt
# Processing test seen dataset (test_random_split.json)
python ${DIR}/tasks/msdp/preprocessing.py \
--func process_wow_dataset \
--raw_file ${WOW_DATA_FOLDER}/test_random_split.json \
--processed_file ${WOW_DATA_FOLDER}/testseen_processed.txt \
--knwl_ref_file ${WOW_DATA_FOLDER}/output_testseen_knowledge_reference.txt \
--resp_ref_file ${WOW_DATA_FOLDER}/output_testseen_response_reference.txt
# processing test unseen dataset (test_topic_split.json)
python ${DIR}/tasks/msdp/preprocessing.py \
--func process_wow_dataset \
--raw_file ${WOW_DATA_FOLDER}/test_topic_split.json \
--processed_file ${WOW_DATA_FOLDER}/testunseen_processed.txt \
--knwl_ref_file ${WOW_DATA_FOLDER}/output_testunseen_knowledge_reference.txt \
--resp_ref_file ${WOW_DATA_FOLDER}/output_testunseen_response_reference.txt
# We provide the following script to process the raw data from Wizard of Internet
# Processing the test dataset (test.jsonl)
python ${DIR}/tasks/msdp/preprocessing.py \
--func process_woi_dataset \
--raw_file ${WOI_DATA_FOLDER}/test.jsonl \
--processed_file ${WOI_DATA_FOLDER}/test_processed.txt \
--knwl_ref_file ${WOI_DATA_FOLDER}/output_test_knowledge_reference.txt \
--resp_ref_file ${WOI_DATA_FOLDER}/output_test_response_reference.txt
# Get the knowledge generation prompts for the each test dataset in WoW and WoI
MODEL_FILE=<PATH_OF_THE_FINETUNED_DPR_MODEL>
# WoW test seen
python ${DIR}/tasks/msdp/preprocessing.py \
--func get_knwl_gen_prompts \
--test_file ${WOW_DATA_FOLDER}/testseen_processed.txt \
--train_file ${WOW_DATA_FOLDER}/train_processed.txt \
--model_file ${MODEL_FILE} \
--processed_file ${WOW_DATA_FOLDER}/output_testseen_knowledge_prompts.json \
--data_type wow_seen
# WoW test unseen
python ${DIR}/tasks/msdp/preprocessing.py \
--func get_knwl_gen_prompts \
--test_file ${WOW_DATA_FOLDER}/testunseen_processed.txt \
--train_file ${WOW_DATA_FOLDER}/train_processed.txt \
--model_file ${MODEL_FILE} \
--processed_file ${WOW_DATA_FOLDER}/output_testunseen_knowledge_prompts.json \
--data_type wow_unseen
# WoI
python ${DIR}/tasks/msdp/preprocessing.py \
--func get_knwl_gen_prompts \
--test_file ${WOI_DATA_FOLDER}/test_processed.txt \
--train_file ${WOW_DATA_FOLDER}/train_processed.txt \
--model_file ${MODEL_FILE} \
--processed_file ${WOI_DATA_FOLDER}/output_test_knowledge_prompts.json \
--data_type woi
# Get the response generation prompts (can be applied for all the test datasets)
python ${DIR}/tasks/msdp/preprocessing.py \
--func get_resp_gen_prompts \
--train_file ${WOW_DATA_FOLDER}/train_processed.txt \
--processed_file ${WOW_DATA_FOLDER}/output_response_prompts.txt
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