unsloth/unsloth/models/rl_replacements.py
2025-06-25 19:51:43 -07:00

495 lines
23 KiB
Python

# Copyright 2023-present Daniel Han-Chen & the Unsloth team. 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.
__all__ = [
"RL_EXTRA_ARGS",
"RL_FUNCTIONS",
"RL_PRE_ITEMS",
"RL_CONFIG_CHANGES",
"RL_METRICS_CHANGES",
]
import os
import re
import torch
import inspect
from collections import defaultdict
from unsloth_zoo.rl_replacements import RL_REPLACEMENTS
RL_EXTRA_ARGS = defaultdict(list)
RL_FUNCTIONS = defaultdict(list)
RL_PRE_ITEMS = defaultdict(list)
RL_CONFIG_CHANGES = defaultdict(list)
RL_METRICS_CHANGES = defaultdict(list)
torch_compile_options = {
"epilogue_fusion" : True,
"max_autotune" : True,
"shape_padding" : True,
"trace.enabled" : False,
"triton.cudagraphs" : False,
}
# Check untrained tokens
def sft_trainer_fix_untrained_tokens(call_args, extra_args):
if "model" in call_args and "train_dataset" in call_args:
fix_tokenizer = \
"IGNORED_TOKENIZER_NAMES = os.environ.get('UNSLOTH_IGNORED_TOKENIZER_NAMES', '').split('\\n')\n"\
"from unsloth_zoo.tokenizer_utils import fix_untrained_tokens\n"\
"from unsloth_zoo.training_utils import fix_zero_training_loss\n"\
"if 'tokenizer' not in locals(): tokenizer = processing_class\n"\
"fix_untrained_tokens(model, tokenizer, train_dataset, IGNORED_TOKENIZER_NAMES, eps = 1e-16)\n"\
"fix_zero_training_loss(model, tokenizer, train_dataset)\n"
return fix_tokenizer
return ""
pass
RL_EXTRA_ARGS["sft_trainer"].append(sft_trainer_fix_untrained_tokens)
# Remove DPO columns which might randomnly be tokenized
def dpo_trainer_fix_columns(call_args, extra_args):
if "model" in call_args and "train_dataset" in call_args:
fix_dpo = \
"if hasattr(train_dataset, 'column_names'):\n"\
" column_names = set(train_dataset.column_names)\n"\
" check = ['chosen', 'rejected', 'prompt', 'chosen_input_ids', 'chosen_attention_mask',\n"\
" 'chosen_labels', 'rejected_input_ids', 'rejected_attention_mask', 'rejected_labels',\n"\
" 'prompt_input_ids', 'prompt_attention_mask']\n"\
" if all(x in column_names for x in check):\n"\
" train_dataset = train_dataset.remove_columns(['chosen', 'rejected', 'prompt'])\n"\
" del check, column_names\n"
return fix_dpo
return ""
pass
RL_EXTRA_ARGS["dpo_trainer"].append(dpo_trainer_fix_columns)
# Fix tokenizer double BOS
def sft_trainer_prepare_dataset(function_name, function):
if function_name != "_prepare_non_packed_dataloader" and \
function_name != "_prepare_dataset": return function
fast_sft_prepare_dataset = RL_REPLACEMENTS.get("sft_prepare_dataset", None)
if fast_sft_prepare_dataset is not None:
params = inspect.signature(fast_sft_prepare_dataset).parameters.keys()
params = ".*?".join(params)
matched = re.match(
r"[\s]{0,}def _prepare_dataset\(.*?" + params + r".*?\)",
function,
flags = re.MULTILINE | re.DOTALL,
)
if matched:
# Use fast version!
function = inspect.getsource(fast_sft_prepare_dataset)
function = function.split("\n")
function = "\n".join(" "*4 + x for x in function)
function = function.replace("def sft_prepare_dataset", "def _prepare_dataset")
return function
pass
pass
check_text = \
"if 'skip_prepare_dataset' in locals() and skip_prepare_dataset:\n"\
" return dataset\n"\
"if 'tokenizer' not in locals(): tokenizer = processing_class\n"\
"if 'formatting_func' not in locals(): raise RuntimeError('Unsloth: Please file a bug report - `formatting_func` does not exist!')\n"\
"if 'dataset_text_field' not in locals() and 'args' in locals(): dataset_text_field = args.dataset_text_field\n"\
"if 'dataset_text_field' not in locals(): raise RuntimeError('Unsloth: Please file a bug report - `dataset_text_field` does not exist!')\n"\
"test_text = dataset[0][dataset_text_field] if (formatting_func is None and dataset_text_field is not None) else formatting_func(dataset[0])[0]\n"\
"chat_template = getattr(tokenizer, 'chat_template', None)\n"\
"chat_template = '' if chat_template is None else chat_template\n"\
"has_bos_token_already = (test_text.startswith(tokenizer.bos_token) or tokenizer.bos_token in chat_template) "\
"if getattr(tokenizer, 'bos_token', None) is not None else False\n"\
"if 'add_special_tokens' not in locals() and has_bos_token_already:\n"\
" from functools import partial\n"\
" tokenizer_call = tokenizer.__call__\n"\
" tokenizer.__call__ = partial(tokenizer_call, add_special_tokens = False)\n"\
" processing_class = tokenizer\n"\
"else:\n"\
" tokenizer_call = None\n"\
" add_special_tokens = False if has_bos_token_already else locals().get('add_special_tokens', False)\n"
check_text = check_text.split("\n")
check_text = "\n".join(" "*8 + x for x in check_text)
check_text = check_text.rstrip() + "\n"
# .*? matches first match. .+? matches final match.
replacer = re.findall(
r"def " + function_name + r"\(.*?\).*?\:\n",
function,
flags = re.MULTILINE | re.DOTALL,
)
if len(replacer) != 0:
replacer = replacer[0]
function = function.replace(replacer, replacer + check_text)
pass
# Return tokenizer's original state
return_state = "if tokenizer_call is not None: tokenizer.__call__ = tokenizer_call\n"
function = re.sub(
r"\n([ ]{4,})(return .*?[\s]{0,})$",
rf"\1{return_state}\1\2",
function,
)
return function
pass
RL_FUNCTIONS["sft_trainer"].append(sft_trainer_prepare_dataset)
# Ignore mean_token_accuracy since it needs logits
# We override it directly with our version
def sft_trainer_compute_loss(function_name, function):
if function_name != "compute_loss": return function
def compute_loss(self, model, inputs, return_outputs = False, num_items_in_batch = None):
outputs = super().compute_loss(
model,
inputs,
return_outputs = return_outputs,
num_items_in_batch = num_items_in_batch,
)
return outputs
pass
function = inspect.getsource(compute_loss)
return function
pass
RL_FUNCTIONS["sft_trainer"].append(sft_trainer_compute_loss)
# Autocast precision for GRPO
def grpo_trainer__prepare_inputs(function_name, function):
if function_name != "_prepare_inputs": return function
import re
# Try to find the function signature and insert after it
# This matches the function signature and any decorators/comments, then finds the first non-empty line after the signature
pattern = r"(def _prepare_inputs\s*\([^\)]*\)\s*(->\s*[^:]+)?\s*:\s*\n)"
match = re.search(pattern, function)
if match:
sig_end = match.end(1)
rest = function[sig_end:]
rest = re.sub(r"^[ \t]*self\.llm\.wake_up\(\)\s*\n", "", rest)
rest = re.sub(r"^[ \t]*torch\.cuda\.empty_cache\(\)\s*\n", "", rest)
rest = re.sub(r"^[ \t]*free, total = torch.cuda.mem_get_info\(\)\s*\n", "", rest)
rest = re.sub(r"^[ \t]*print\(f?\".*cuda.*\"\)\s*\n", "", rest)
insert = (
" if hasattr(self, 'llm'):\n"
" if getattr(self.llm.llm_engine.vllm_config.model_config, 'enable_sleep_mode', False):\n"
" self.llm.wake_up()\n"
)
function = function[:sig_end] + insert + rest
else:
pattern2 = r"(def _prepare_inputs\(.*?\):\n(?:[ ]+#[^\n]*\n)+)"
match2 = re.search(pattern2, function, flags=re.DOTALL)
if match2:
header_and_comments = match2.group(1)
rest = function[len(header_and_comments):]
rest = re.sub(r"^[ \t]*self\.llm\.wake_up\(\)\s*\n", "", rest)
rest = re.sub(r"^[ \t]*torch\.cuda\.empty_cache\(\)\s*\n", "", rest)
rest = re.sub(r"^[ \t]*free, total = torch.cuda.mem_get_info\(\)\s*\n", "", rest)
rest = re.sub(r"^[ \t]*print\(f?\".*cuda.*\"\)\s*\n", "", rest)
insert = (
" if (hasattr(self, 'llm'):\n"
" if getattr(self.llm.llm_engine.vllm_config.model_config, 'enable_sleep_mode', False):\n"
" self.llm.wake_up()\n"
)
function = header_and_comments + insert + rest
# Add mixed precision training
function = function.replace(
"with torch.inference_mode():",
"with torch.inference_mode(), "\
"torch.amp.autocast(device_type = 'cuda', "\
"dtype = ((torch.float16 if os.environ.get('ACCELERATE_MIXED_PRECISION', 'fp16') == 'fp16' else torch.bfloat16) "\
"if not torch.is_autocast_enabled('cuda') else nullcontext())"\
"if os.environ.get('UNSLOTH_FORCE_FLOAT32', '0') == '0' else torch.float16):",
)
function = function.replace(
"self.accelerator.unwrap_model(self.model)",
"self.accelerator.unwrap_model(self.model, keep_fp32_wrapper = False)",
)
sleep_and_cache = (
"if hasattr(self, 'llm'):\n"
" if getattr(self.llm.llm_engine.vllm_config.model_config, 'enable_sleep_mode', False):\n"
" self.llm.sleep(os.environ.get('VLLM_SLEEP_MODE', 1))\n"
" "
)
if re.search(r"\n\s*return ", function):
function = re.sub(r"(\n\s*)return ", f"\\1{sleep_and_cache}return ", function, count=1)
else:
function = function.rstrip() + "\n " + sleep_and_cache
return function
pass
RL_FUNCTIONS["grpo_trainer"].append(grpo_trainer__prepare_inputs)
# Remove _move_model_to_vllm
def grpo_trainer__move_model_to_vllm(function_name, function):
if function_name != "_move_model_to_vllm": return function
def _move_model_to_vllm(self, *args, **kwargs): return None
function = inspect.getsource(_move_model_to_vllm)
return function
pass
RL_FUNCTIONS["grpo_trainer"].append(grpo_trainer__move_model_to_vllm)
# Edit _get_per_token_logps to handle mixed precision
def grpo_trainer__get_per_token_logps(function_name, function):
if function_name != "_get_per_token_logps": return function
def _get_per_token_logps(self, model, input_ids, attention_mask, logits_to_keep, calc_logprob_flag = None):
if os.environ.get('UNSLOTH_USE_NEW_MODEL', '0') == '0' and not calc_logprob_flag:
return None # Unsloth efficient GRPO
# Otherwise, calculate normally:
if not hasattr(self, '_autocast_dtype'):
self._autocast_dtype = torch.float16 if os.environ.get('ACCELERATE_MIXED_PRECISION', 'fp16') == 'fp16' else torch.bfloat16
if os.environ.get('UNSLOTH_FORCE_FLOAT32', '0') == '1': self._autocast_dtype = torch.float16
os.environ["UNSLOTH_RETURN_HIDDEN_STATES"] = "1"
with torch.amp.autocast(device_type = 'cuda', dtype = self._autocast_dtype):
# We add 1 to `logits_to_keep` because the last logits of the sequence is later excluded
print("input_ids Unsloth 264", input_ids.shape)
print("logits_to_keep Unsloth 264", logits_to_keep)
hidden_states = model(input_ids=input_ids, attention_mask=attention_mask, logits_to_keep=logits_to_keep + 1).logits
print("hidden_states Unsloth 264", hidden_states.shape)
#logits = logits[:, :-1, :] # (B, L-1, V), exclude the last logit: it corresponds to the next token pred
return hidden_states
# input_ids = input_ids[:, -logits_to_keep:]
# For transformers<=4.48, logits_to_keep argument isn't supported, so here we drop logits ourselves.
# See https://github.com/huggingface/trl/issues/2770
# logits = logits[:, -logits_to_keep:]
# return logits
# See https://huggingface.co/blog/the_n_implementation_details_of_rlhf_with_ppo#policy-training-implementation-details
# logits = logits / self.temperature
# logps = selective_log_softmax(logits, input_ids)
# row_indices, col_indices = torch.where(logps < -20)
# # Method 1: Check if tensors have elements
# if len(row_indices) > 0 and len(col_indices) > 0:
# breakpoint() # Breakpoint triggered here
# print("Found high values!")
# return logps # compute logprobs for the input tokens
pass
pass
function = inspect.getsource(_get_per_token_logps)
return function
pass
RL_FUNCTIONS["grpo_trainer"].append(grpo_trainer__get_per_token_logps)
grpo_compute_loss = RL_REPLACEMENTS["grpo_compute_loss"]
grpo_compute_loss_slow = RL_REPLACEMENTS["grpo_compute_loss_slow"]
UnslothEfficientGRPO = RL_REPLACEMENTS["UnslothEfficientGRPO"]
grpo_accumulated_loss = RL_REPLACEMENTS["grpo_accumulated_loss"]
RL_PRE_ITEMS["grpo_trainer"].append(inspect.getsource(grpo_compute_loss))
RL_PRE_ITEMS["grpo_trainer"].append(inspect.getsource(UnslothEfficientGRPO))
RL_PRE_ITEMS["grpo_trainer"].append(inspect.getsource(grpo_accumulated_loss))
RL_PRE_ITEMS["grpo_trainer"].append(grpo_compute_loss_slow)
# Edit _get_per_token_logps to handle mixed precision
def grpo_trainer_compute_loss(function_name, function):
if function_name != "compute_loss": return function
def compute_loss(self, model, inputs, return_outputs = False, num_items_in_batch = None):
if return_outputs:
raise ValueError("The GRPOTrainer does not support returning outputs")
# Compute the per-token log probabilities for the model
prompt_ids, prompt_mask = inputs["prompt_ids"], inputs["prompt_mask"]
completion_ids, completion_mask = inputs["completion_ids"], inputs["completion_mask"]
input_ids = torch.cat([prompt_ids, completion_ids], dim=1)
bsz, qlen = input_ids.shape
attention_mask = torch.cat([prompt_mask, completion_mask], dim=1)
# attention_mask = None
logits_to_keep = completion_ids.size(1) # we only need to compute the logits for the completion tokens
_input_ids = input_ids
_logits_to_keep = logits_to_keep
print("prompt_mask Unsloth 320", prompt_mask.shape)
print("completion_mask Unsloth 320", completion_mask.shape)
print("input_ids Unsloth 320", input_ids.shape)
print("logits_to_keep Unsloth 320", logits_to_keep)
per_token_logps = self._get_per_token_logps(model, input_ids, attention_mask, logits_to_keep)
if per_token_logps is not None:
print("per_token_logps Unsloth 320", per_token_logps.shape)
# Compute the KL divergence between the model and the reference model
# _prepare_inputs doesn't return reference log probs anymore. We need to calculate it ourselves.
# https://github.com/huggingface/trl/blob/05bc43e960396581e458195b8388efe6b82cae1f/trl/trainer/grpo_trainer.py#L1328
if self.beta != 0.0:
with torch.inference_mode(), model.disable_adapter():
ref_per_token_logps = self._get_per_token_logps(model, input_ids, attention_mask, logits_to_keep)
print("ref_per_token_logps Unsloth 320", ref_per_token_logps.shape)
else:
ref_per_token_logps = None
# per_token_kl = torch.exp(ref_per_token_logps - per_token_logps) - (ref_per_token_logps - per_token_logps) - 1
# x - x.detach() allows for preserving gradients from x
advantages = inputs["advantages"]
print("advantages Unsloth 320", advantages.shape)
# per_token_loss = torch.exp(per_token_logps - per_token_logps.detach()) * advantages.unsqueeze(1)
# per_token_loss = -(per_token_loss - self.beta * per_token_kl)
# loss = ((per_token_loss * completion_mask).sum(dim=1) / completion_mask.sum(dim=1)).mean()
old_hidden_states = inputs.get("old_per_token_logps", None)
if old_hidden_states is not None:
print("old_hidden_states Unsloth 320", old_hidden_states.shape)
print("input_ids Unsloth 320", input_ids.shape)
print("logits_to_keep Unsloth 320", logits_to_keep)
input_ids = input_ids[:, -logits_to_keep:]
print("input_ids Unsloth 320", input_ids.shape)
# Get logit softcapping and logit scale
logit_softcapping = getattr(model.config, "final_logit_softcapping", 0) # Gemma
if logit_softcapping is None: logit_softcapping = 0
logit_scale_multiply = getattr(model.config, "logit_scale", 0) # Cohere
if logit_scale_multiply is None: logit_scale_multiply = 0
logit_scale_divide = getattr(model.config, "logits_scaling", 0) # Granite
if logit_scale_divide is None: logit_scale_divide = 0
if per_token_logps is not None:
if ref_per_token_logps is not None:
print("ref_per_token_logps Unsloth 320", ref_per_token_logps.shape)
ref_per_token_logps = ref_per_token_logps[:, :-1, :] # (B, L-1, V), exclude the last logit: it corresponds to the next token pred
print("ref_per_token_logps Unsloth 320", ref_per_token_logps.shape)
print("per_token_logps Unsloth 320", per_token_logps.shape)
per_token_logps = per_token_logps[:, :-1, :] # (B, L-1, V), exclude the last logit: it corresponds to the next token pred
print("per_token_logps Unsloth 320", per_token_logps.shape)
loss, completion_length, mean_kl = grpo_compute_loss_slow(
ref_per_token_logps,
per_token_logps,
old_hidden_states,
input_ids,
completion_mask,
self.beta,
advantages,
loss_type = self.args.loss_type,
epsilon_low = self.epsilon_low,
epsilon_high = self.epsilon_high,
max_completion_length = self.args.max_completion_length,
delta = self.args.delta,
temperature = self.args.temperature,
logit_softcapping = logit_softcapping,
logit_scale_multiply = logit_scale_multiply,
logit_scale_divide = logit_scale_divide,
)
else:
if hasattr(self.args, "loss_type"):
loss, completion_length, mean_kl = grpo_accumulated_loss(
self,
_input_ids,
logits_to_keep,
completion_mask,
advantages,
old_hidden_states,
n_chunks = self.args.unsloth_num_chunks,
loss_type = self.args.loss_type,
epsilon_low = self.epsilon_low,
epsilon_high = self.epsilon_high,
max_completion_length = self.args.max_completion_length,
delta = self.args.delta,
temperature = self.args.temperature,
logit_softcapping = logit_softcapping,
logit_scale_multiply = logit_scale_multiply,
logit_scale_divide = logit_scale_divide,
)
else:
# to ensure backwards compatibility with trl 0.15.2 and maybe even 0.17
loss, completion_length, mean_kl = grpo_accumulated_loss(
self,
_input_ids,
logits_to_keep,
completion_mask,
advantages,
old_hidden_states,
n_chunks = self.args.unsloth_num_chunks,
temperature = self.args.temperature,
logit_softcapping = logit_softcapping,
logit_scale_multiply = logit_scale_multiply,
logit_scale_divide = logit_scale_divide,
)
# Log the metrics
# completion_length = self.accelerator.gather_for_metrics(completion_mask.sum(1)).float().mean().item()
# mean_kl = ((per_token_kl * completion_mask).sum(dim=1) / completion_mask.sum(dim=1)).mean()
# self._metrics["kl"].append(self.accelerator.gather_for_metrics(mean_kl).mean().item())
if "train" in self._metrics:
mode = "eval" if self.control.should_evaluate else "train"
self._metrics[mode]["completion_length"].append(completion_length.item())
self._metrics[mode]["kl"].append(mean_kl.item())
else:
self._metrics["completion_length"].append(completion_length.item())
self._metrics["kl"].append(mean_kl.item())
return loss
pass
function = inspect.getsource(compute_loss)
return function
pass
RL_FUNCTIONS["grpo_trainer"].append(grpo_trainer_compute_loss)
# https://github.com/huggingface/trl/blob/main/trl/trainer/grpo_trainer.py#L356
# TRL warns if batch size is not a multiple of num_generations -> fix this.
def grpo_trainer_fix_batch_size(RLTrainer_source, RLConfig_source):
if "divisible by the number of generations" not in RLTrainer_source: return ""
if "num_generations" not in RLConfig_source: return ""
check_batch_size = \
"div = per_device_train_batch_size // num_generations\n"\
"if div * num_generations != per_device_train_batch_size:\n"\
" print('Unsloth: We now expect `per_device_train_batch_size` to be a multiple of `num_generations`.\\n"\
"We will change the batch size of ' + str(per_device_train_batch_size) + ' to the `num_generations` of ' + str(num_generations))\n"\
" per_device_train_batch_size = num_generations\n"
return check_batch_size
pass
RL_CONFIG_CHANGES["grpo_trainer"].append(grpo_trainer_fix_batch_size)
# Add other reward function names
def grpo_trainer_metrics(RLTrainer_source, RLConfig_source):
if "reward_funcs" not in RLTrainer_source: return ""
# For new TRL we have /mean and /std
use_mean = "rewards/{reward_func_name}/mean" in RLTrainer_source
use_std = "rewards/{reward_func_name}/std" in RLTrainer_source
if not use_mean:
use_normal = "rewards/{reward_func_name}" in RLTrainer_source
else:
use_normal = False
pass
log_metrics = \
"if not isinstance(reward_funcs, list): _reward_funcs = [reward_funcs]\n"\
"else: _reward_funcs = reward_funcs\n"\
"for reward_func in _reward_funcs:\n"\
" try:\n"\
" reward_func_name = reward_func.__name__\n"\
f" if {use_mean}:\n"\
" other_metrics.append(f'rewards/{reward_func_name}/mean')\n"\
f" if {use_std}:\n"\
" other_metrics.append(f'rewards/{reward_func_name}/std')\n"\
f" if {use_normal}:\n"\
" other_metrics.append(f'rewards/{reward_func_name}')\n"\
" except: pass\n"
return log_metrics
pass
RL_METRICS_CHANGES["grpo_trainer"].append(grpo_trainer_metrics)