support for custom trainer_cls from config

* FIX: TRL trainer preprocessing step was running in one process

* FIX: max_length and max_prompt_length was not being sent to ORPOTrainer

* FIX: Change ORPO max prompt length to 1/4 of max length, otherwise we get strange behaviour

* FIX: Removed change from a different PR

* FIX: Black fix

* explicitly set max prompt len for orpo config

---------

Co-authored-by: Ali Mosavian <ali.mosavian@kry.se>
Co-authored-by: Wing Lian <wing.lian@gmail.com>

README.md

@@ -34,6 +34,7 @@ Features:
   - [Mac](#mac)
   - [Google Colab](#google-colab)
   - [Launching on public clouds via SkyPilot](#launching-on-public-clouds-via-skypilot)
+  - [Launching on public clouds via dstack](#launching-on-public-clouds-via-dstack)
 - [Dataset](#dataset)
 - [Config](#config)
   - [Train](#train)
@@ -292,6 +293,42 @@ HF_TOKEN=xx sky launch axolotl.yaml --env HF_TOKEN
 HF_TOKEN=xx BUCKET=<unique-name> sky spot launch axolotl-spot.yaml --env HF_TOKEN --env BUCKET
 ```
 
+#### Launching on public clouds via dstack
+To launch on GPU instance (both on-demand and spot instances) on public clouds (GCP, AWS, Azure, Lambda Labs, TensorDock, Vast.ai, and CUDO), you can use [dstack](https://dstack.ai/).
+
+Write a job description in YAML as below:
+
+```yaml
+# dstack.yaml
+type: task
+
+image: winglian/axolotl-cloud:main-20240429-py3.11-cu121-2.2.1
+
+env:
+  - HUGGING_FACE_HUB_TOKEN
+  - WANDB_API_KEY
+
+commands:
+  - accelerate launch -m axolotl.cli.train config.yaml
+
+ports:
+  - 6006
+
+resources:
+  gpu:
+    memory: 24GB..
+    count: 2
+```
+
+then, simply run the job with `dstack run` command. Append `--spot` option if you want spot instance. `dstack run` command will show you the instance with cheapest price across multi cloud services:
+
+```bash
+pip install dstack
+HUGGING_FACE_HUB_TOKEN=xxx WANDB_API_KEY=xxx dstack run . -f dstack.yaml # --spot
+```
+
+For further and fine-grained use cases, please refer to the official [dstack documents](https://dstack.ai/docs/) and the detailed description of [axolotl example](https://github.com/dstackai/dstack/tree/master/examples/fine-tuning/axolotl) on the official repository.
+
 ### Dataset
 
 Axolotl supports a variety of dataset formats.  It is recommended to use a JSONL.  The schema of the JSONL depends upon the task and the prompt template you wish to use.  Instead of a JSONL, you can also use a HuggingFace dataset with columns for each JSONL field.

docs/config.qmd

@@ -138,7 +138,7 @@ test_datasets:
     data_files:
       - /workspace/data/eval.jsonl
 
-# use RL training: 'dpo', 'ipo', 'kto_pair', 'orpo', 'sppo_hard'
+# use RL training: 'dpo', 'ipo', 'kto_pair'
 rl:
 
 # Saves the desired chat template to the tokenizer_config.json for easier inferencing

requirements.txt

@@ -28,7 +28,7 @@ scipy
 scikit-learn==1.2.2
 pynvml
 art
-fschat @ git+https://github.com/lm-sys/FastChat.git@5095615810cf613dba7f27dd155f571fcff976d8
+fschat @ git+https://github.com/lm-sys/FastChat.git@27a05b04a35510afb1d767ae7e5990cbd278f8fe
 gradio==3.50.2
 tensorboard
 
@@ -39,6 +39,6 @@ s3fs
 gcsfs
 # adlfs
 
-trl @ git+https://github.com/huggingface/trl.git@75de236c09bd5846f79c24d9bf371481b0b7582c
+trl==0.8.5
 zstandard==0.22.0
 fastcore

src/axolotl/cli/merge_lora.py

@@ -25,6 +25,8 @@ def do_cli(config: Path = Path("examples/"), **kwargs):
         load_in_8bit=False,
         load_in_4bit=False,
         flash_attention=False,
+        deepspeed=None,
+        fsdp=None,
         **kwargs,
     )
 
@@ -40,6 +42,7 @@ def do_cli(config: Path = Path("examples/"), **kwargs):
     parsed_cfg.flash_attention = False
     parsed_cfg.deepspeed = None
     parsed_cfg.fsdp = None
+    parsed_cfg.fsdp_config = None
 
     do_merge_lora(cfg=parsed_cfg, cli_args=parsed_cli_args)
 

src/axolotl/cli/preprocess.py

@@ -19,7 +19,10 @@ from axolotl.cli import (
 )
 from axolotl.common.cli import PreprocessCliArgs
 from axolotl.common.const import DEFAULT_DATASET_PREPARED_PATH
-from axolotl.prompt_strategies.sharegpt import register_chatml_template
+from axolotl.prompt_strategies.sharegpt import (
+    register_chatml_template,
+    register_llama3_template,
+)
 
 LOG = logging.getLogger("axolotl.cli.preprocess")
 
@@ -36,13 +39,22 @@ def do_cli(config: Union[Path, str] = Path("examples/"), **kwargs):
         return_remaining_strings=True
     )
 
-    if parsed_cfg.chat_template == "chatml" and parsed_cfg.default_system_message:
+    if parsed_cfg.chat_template == "chatml":
-        LOG.info(
+        if parsed_cfg.default_system_message:
-            f"ChatML set. Adding default system message: {parsed_cfg.default_system_message}"
+            LOG.info(
-        )
+                f"ChatML set. Adding default system message: {parsed_cfg.default_system_message}"
-        register_chatml_template(parsed_cfg.default_system_message)
+            )
-    else:
+            register_chatml_template(parsed_cfg.default_system_message)
-        register_chatml_template()
+        else:
+            register_chatml_template()
+    elif parsed_cfg.chat_template == "llama3":
+        if parsed_cfg.default_system_message:
+            LOG.info(
+                f"LLaMA-3 set. Adding default system message: {parsed_cfg.default_system_message}"
+            )
+            register_llama3_template(parsed_cfg.default_system_message)
+        else:
+            register_llama3_template()
 
     if not parsed_cfg.dataset_prepared_path:
         msg = (

src/axolotl/cli/train.py

@@ -19,7 +19,10 @@ from axolotl.cli import (
     print_axolotl_text_art,
 )
 from axolotl.common.cli import TrainerCliArgs
-from axolotl.prompt_strategies.sharegpt import register_chatml_template
+from axolotl.prompt_strategies.sharegpt import (
+    register_chatml_template,
+    register_llama3_template,
+)
 from axolotl.train import train
 
 LOG = logging.getLogger("axolotl.cli.train")
@@ -47,6 +50,14 @@ def do_train(cfg, cli_args) -> Tuple[PreTrainedModel, PreTrainedTokenizer]:
     else:
         register_chatml_template()
 
+    if cfg.chat_template == "llama3" and cfg.default_system_message:
+        LOG.info(
+            f"LLaMA-3 set. Adding default system message: {cfg.default_system_message}"
+        )
+        register_llama3_template(cfg.default_system_message)
+    else:
+        register_llama3_template()
+
     if cfg.rl:  # and cfg.rl != "orpo":
         dataset_meta = load_rl_datasets(cfg=cfg, cli_args=cli_args)
     else:

src/axolotl/core/trainer_builder.py

@@ -30,7 +30,7 @@ from transformers import (
 )
 from transformers.trainer_utils import seed_worker
 from transformers.utils import is_sagemaker_mp_enabled
-from trl import DPOConfig, DPOTrainer, ORPOConfig, ORPOTrainer
+from trl import DPOTrainer, ORPOConfig, ORPOTrainer
 from trl.trainer.utils import pad_to_length
 
 from axolotl.loraplus import create_loraplus_optimizer
@@ -993,6 +993,9 @@ class HFCausalTrainerBuilder(TrainerBuilderBase):
             return ReLoRATrainer
         if self.cfg.model_config_type == "mamba":
             return AxolotlMambaTrainer
+        if self.cfg.custom_trainer_cls:
+            _module, _cls = self.cfg.custom_trainer_cls.rsplit(".", 1)
+            return importlib.import_module(_module, _cls)
         return AxolotlTrainer
 
     def build(self, total_num_steps):
@@ -1526,9 +1529,9 @@ class HFRLTrainerBuilder(TrainerBuilderBase):
         if self.cfg.rl == "orpo":
             training_args_cls = ORPOConfig
             training_args_kwargs["dataset_num_proc"] = self.cfg.dataset_processes
-        elif self.cfg.rl in ["dpo", "ipo", "kto_pair", "sppo_hard"]:
+            training_args_kwargs["max_length"] = self.cfg.sequence_len
-            training_args_cls = DPOConfig
+            if self.cfg.max_prompt_len:
-            training_args_kwargs["dataset_num_proc"] = self.cfg.dataset_processes
+                training_args_kwargs["max_prompt_length"] = self.cfg.max_prompt_len
 
         training_args = training_args_cls(
             per_device_train_batch_size=self.cfg.micro_batch_size,
@@ -1555,8 +1558,6 @@ class HFRLTrainerBuilder(TrainerBuilderBase):
                 dpo_trainer_kwargs["label_smoothing"] = self.cfg.dpo_label_smoothing
         elif self.cfg.rl == "kto_pair":
             dpo_trainer_kwargs["loss_type"] = "kto_pair"
-        elif self.cfg.rl == "sppo_hard":
-            dpo_trainer_kwargs["loss_type"] = "sppo_hard"
         if self.eval_dataset:
             dpo_trainer_kwargs["eval_dataset"] = self.eval_dataset
         if self.cfg.adapter and self.peft_config:
@@ -1565,7 +1566,7 @@ class HFRLTrainerBuilder(TrainerBuilderBase):
             dpo_trainer_kwargs[
                 "precompute_ref_log_probs"
             ] = self.cfg.precompute_ref_log_probs
-        if self.cfg.rl in ["dpo", "ipo", "kto_pair", "sppo_hard"]:
+        if self.cfg.rl in ["dpo", "ipo", "kto_pair"]:
             trainer_cls = AxolotlDPOTrainer
             dpo_trainer_kwargs["beta"] = self.cfg.dpo_beta or 0.1
             trainer_cls_args = [self.model, self.model_ref]

src/axolotl/monkeypatch/fastchat_conversation_turns.py

@@ -123,6 +123,17 @@ def get_turns(  # pylint: disable=too-many-return-statements
             else:
                 yield role, ""
         return
+    if self.sep_style == SeparatorStyle.LLAMA3:
+        if self.system_message:
+            # For llama3, the system message is NOT incorporated into the first human instruction
+            # All messages follow <|start_header_id|>' + role + '<|end_header_id|>\n\n'+ message + '<|eot_id|>
+            yield "", system_prompt
+        for i, (role, message) in enumerate(self.messages):
+            if message:
+                yield f"<|start_header_id|>{role}<|end_header_id|>\n\n", f"{message.strip()}<|eot_id|>"
+            else:
+                yield f"<|start_header_id|>{role}<|end_header_id|>\n\n", ""
+        return
     if self.sep_style == SeparatorStyle.GEMMA:
         if self.system_message:
             raise ValueError("Gemma chat template does not support system messages")

src/axolotl/prompt_strategies/dpo/llama3.py

@@ -0,0 +1,133 @@
+"""
+DPO strategies for llama-3 chat template
+"""
+
+
+def argilla(
+    cfg,
+    **kwargs,
+):  # pylint: disable=possibly-unused-variable,unused-argument
+    def transform_fn(sample):
+        if "system" in sample and sample["system"]:
+            sample["prompt"] = (
+                f"<|start_header_id|>system<|end_header_id|>\n\n{sample['system']}<|eot_id|>"
+                f"<|start_header_id|>user<|end_header_id|>\n\n{sample['instruction']}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
+            )
+        else:
+            sample[
+                "prompt"
+            ] = f"<|start_header_id|>user<|end_header_id|>\n\n{sample['instruction']}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
+        sample["chosen"] = f"{sample['chosen_response']}<|eot_id|>"
+        sample["rejected"] = f"{sample['rejected_response']}<|eot_id|>"
+        return sample
+
+    return transform_fn
+
+
+def argilla_chat(
+    cfg,
+    **kwargs,
+):  # pylint: disable=possibly-unused-variable,unused-argument
+    """
+    for argilla/dpo-mix-7k conversations
+    """
+
+    def transform_fn(sample):
+        sample[
+            "prompt"
+        ] = f"<|start_header_id|>user<|end_header_id|>\n\n{sample['chosen'][0]['content']}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
+        sample["chosen"] = f"{sample['chosen'][1]['content']}<|eot_id|>"
+        sample["rejected"] = f"{sample['rejected'][1]['content']}<|eot_id|>"
+        return sample
+
+    return transform_fn
+
+
+def icr(
+    cfg,
+    **kwargs,
+):  # pylint: disable=possibly-unused-variable,unused-argument
+    """
+    chatml transforms for datasets with system, input, chosen, rejected
+    ex. https://huggingface.co/datasets/argilla/distilabel-intel-orca-dpo-pairs
+    """
+
+    def transform_fn(sample):
+        if "system" in sample and sample["system"]:
+            sample["prompt"] = (
+                f"<|start_header_id|>system<|end_header_id|>\n\n{sample['system']}<|eot_id|>"
+                f"<|start_header_id|>user<|end_header_id|>\n\n{sample['input']}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
+            )
+        else:
+            sample[
+                "prompt"
+            ] = f"<|start_header_id|>user<|end_header_id|>\n\n{sample['input']}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
+        sample["chosen"] = f"{sample['chosen']}<|eot_id|>"
+        sample["rejected"] = f"{sample['rejected']}<|eot_id|>"
+        return sample
+
+    return transform_fn
+
+
+def intel(cfg, **kwargs):  # pylint: disable=possibly-unused-variable,unused-argument
+    """
+    For Intel Orca DPO Pairs
+    """
+
+    def transform_fn(sample):
+        if "system" in sample and sample["system"]:
+            sample["prompt"] = (
+                f"<|start_header_id|>system<|end_header_id|>\n\n{sample['system']}<|eot_id|>"
+                f"<|start_header_id|>user<|end_header_id|>\n\n{sample['question']}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
+            )
+        else:
+            sample[
+                "prompt"
+            ] = f"<|start_header_id|>user<|end_header_id|>\n\n{sample['question']}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
+        sample["chosen"] = f"{sample['chosen']}<|eot_id|>"
+        sample["rejected"] = f"{sample['rejected']}<|eot_id|>"
+        return sample
+
+    return transform_fn
+
+
+def prompt_pairs(
+    cfg, **kwargs
+):  # pylint: disable=possibly-unused-variable,unused-argument
+    def transform_fn(sample):
+        if "system" in sample and sample["system"]:
+            sample["prompt"] = (
+                f"<|start_header_id|>system<|end_header_id|>\n\n{sample['system']}<|eot_id|>"
+                f"<|start_header_id|>user<|end_header_id|>\n\n{sample['prompt']}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
+            )
+        else:
+            sample[
+                "prompt"
+            ] = f"<|start_header_id|>user<|end_header_id|>\n\n{sample['prompt']}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
+        sample["chosen"] = f"{sample['chosen']}<|eot_id|>"
+        sample["rejected"] = f"{sample['rejected']}<|eot_id|>"
+        return sample
+
+    return transform_fn
+
+
+def ultra(cfg, **kwargs):  # pylint: disable=possibly-unused-variable,unused-argument
+    """
+    for ultrafeedback binarized conversations
+    """
+
+    def transform_fn(sample):
+        if "system" in sample and sample["system"]:
+            sample["prompt"] = (
+                f"<|start_header_id|>system<|end_header_id|>\n\n{sample['system']}<|eot_id|>"
+                f"<|start_header_id|>user<|end_header_id|>\n\n{sample['prompt']}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
+            )
+        else:
+            sample[
+                "prompt"
+            ] = f"<|start_header_id|>user<|end_header_id|>\n\n{sample['prompt']}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
+        sample["chosen"] = f"{sample['chosen'][1]['content']}<|eot_id|>"
+        sample["rejected"] = f"{sample['rejected'][1]['content']}<|eot_id|>"
+        return sample
+
+    return transform_fn

src/axolotl/prompt_strategies/dpo/mistral.py

@@ -1,30 +0,0 @@
-"""
-DPO strategies for mistral instruct
-"""
-
-
-def prompt_pairs(cfg):  # pylint: disable=possibly-unused-variable,unused-argument
-    def transform_fn(sample):
-        sample["prompt"] = f"[INST]{sample['prompt']}[/INST]"
-        sample["chosen"] = f"{sample['chosen']}"
-        sample["rejected"] = f"{sample['rejected']}"
-        return sample
-
-    return transform_fn
-
-
-def argilla_chat(
-    cfg,
-    **kwargs,
-):  # pylint: disable=possibly-unused-variable,unused-argument
-    """
-    for argilla/dpo-mix-7k conversations
-    """
-
-    def transform_fn(sample):
-        sample["prompt"] = f"[INST] {sample['chosen'][0]['content']} [/INST]"
-        sample["chosen"] = f"{sample['chosen'][1]['content']}</s>"
-        sample["rejected"] = f"{sample['rejected'][1]['content']}</s>"
-        return sample
-
-    return transform_fn

src/axolotl/prompt_strategies/sharegpt.py

@@ -1,7 +1,7 @@
 """Module containing the SimpleShareGPTPromptTokenizingStrategy class"""
 
 import logging
-from typing import Any, Dict, Optional
+from typing import Any, Dict, Optional, Type
 
 from fastchat.conversation import Conversation, SeparatorStyle, register_conv_template
 
@@ -22,7 +22,7 @@ def register_chatml_template(system_message=None):
             name="chatml",
             system_template="<|im_start|>system\n{system_message}",
             system_message=system_message,
-            roles=["<|im_start|>user", "<|im_start|>assistant"],
+            roles=("<|im_start|>user", "<|im_start|>assistant"),
             sep_style=SeparatorStyle.CHATML,
             sep="<|im_end|>",
         )
@@ -32,83 +32,63 @@ def register_chatml_template(system_message=None):
             name="chatml_glaive",
             system_template="<|im_start|>system\n{system_message}",
             system_message=system_message,
-            roles=["<|im_start|>user", "<|im_start|>assistant", "<|im_start|>tool"],
+            roles=("<|im_start|>user", "<|im_start|>assistant", "<|im_start|>tool"),
             sep_style=SeparatorStyle.CHATML,
             sep="<|im_end|>",
         )
     )
 
 
-def load(tokenizer, cfg, ds_cfg: Optional[Dict[str, Any]] = None):
+def register_llama3_template(system_message=None):
-    conversation = (
+    system_message = system_message or "You are a helpful assistant."
-        ds_cfg["conversation"] if ds_cfg and "conversation" in ds_cfg else None
+    register_conv_template(
-    )
+        Conversation(
-    field_human = ds_cfg["field_human"] if ds_cfg and "field_human" in ds_cfg else None
+            name="llama3",
-    field_model = ds_cfg["field_model"] if ds_cfg and "field_model" in ds_cfg else None
+            system_template="<|start_header_id|>system<|end_header_id|>\n\n{system_message}<|eot_id|>",
-    roles = ds_cfg["roles"].to_dict() if ds_cfg and "roles" in ds_cfg else None
+            system_message=system_message,
-    strategy = SimpleShareGPTPromptTokenizingStrategy(
+            roles=("user", "assistant"),
-        ShareGPTPrompterV2(
+            sep_style=SeparatorStyle.LLAMA3,
-            conversation=conversation,
+            sep="",
-            role_key_model=field_model,
+            stop_str="<|eot_id|>",
-            role_key_human=field_human,
+            stop_token_ids=[128001, 128009],
-            roles=roles,
+        )
-        ),
-        tokenizer,
-        cfg.train_on_inputs,
-        cfg.sequence_len,
-    )
-    if ds_cfg and "strict" in ds_cfg:
-        strategy.strict = ds_cfg["strict"]
-    return strategy
-
-
-def load_ultrachat(tokenizer, cfg, ds_cfg: Optional[Dict[str, Any]] = None):
-    conversation = (
-        ds_cfg["conversation"] if ds_cfg and "conversation" in ds_cfg else None
-    )
-    strategy = UltrachatShareGPTPromptTokenizingStrategy(
-        ShareGPTPrompterV2(
-            conversation=conversation,
-        ),
-        tokenizer,
-        cfg.train_on_inputs,
-        cfg.sequence_len,
-    )
-    if ds_cfg and "strict" in ds_cfg:
-        strategy.strict = ds_cfg["strict"]
-    return strategy
-
-
-def load_role(tokenizer, cfg):
-    return SimpleRoleShareGPTPromptTokenizingStrategy(
-        ShareGPTPrompterV2(),
-        tokenizer,
-        cfg.train_on_inputs,
-        cfg.sequence_len,
     )
 
 
-def load_guanaco(tokenizer, cfg):
+def build_loader(
-    return GuanacoShareGPTPromptTokenizingStrategy(
+    tokenization_strategy_cls: Type["ShareGPTPromptTokenizingStrategy"],
-        ShareGPTPrompterV2(),
+    prompter_cls: Type["ShareGPTPrompterV2"],
-        tokenizer,
+    default_conversation: Optional[str] = None,
-        cfg.train_on_inputs,
+):
-        cfg.sequence_len,
+    def _load(tokenizer, cfg, ds_cfg: Optional[Dict[str, Any]] = None):
-    )
+        conversation = (
+            ds_cfg["conversation"]
+            if ds_cfg and "conversation" in ds_cfg
+            else default_conversation
+        )
+        field_human = (
+            ds_cfg["field_human"] if ds_cfg and "field_human" in ds_cfg else None
+        )
+        field_model = (
+            ds_cfg["field_model"] if ds_cfg and "field_model" in ds_cfg else None
+        )
+        roles = ds_cfg["roles"].to_dict() if ds_cfg and "roles" in ds_cfg else None
+        strategy = tokenization_strategy_cls(
+            prompter_cls(
+                conversation=conversation,
+                role_key_model=field_model,
+                role_key_human=field_human,
+                roles=roles,
+            ),
+            tokenizer,
+            cfg.train_on_inputs,
+            cfg.sequence_len,
+        )
+        if ds_cfg and "strict" in ds_cfg and hasattr(strategy, "strict"):
+            strategy.strict = ds_cfg["strict"]
+        return strategy
 
-
+    return _load
-def load_glaive(tokenizer, cfg, ds_cfg: Optional[Dict[str, Any]] = None):
-    conversation = (
-        ds_cfg["conversation"]
-        if ds_cfg and "conversation" in ds_cfg
-        else "chatml_glaive"
-    )
-    return GlaiveShareGPTPromptTokenizingStrategy(
-        ShareGPTPrompterV2(conversation=conversation),
-        tokenizer,
-        cfg.train_on_inputs,
-        cfg.sequence_len,
-    )
 
 
 class SimpleShareGPTPromptTokenizingStrategy(ShareGPTPromptTokenizingStrategy):
@@ -158,7 +138,9 @@ class SimpleShareGPTPromptTokenizingStrategy(ShareGPTPromptTokenizingStrategy):
         return turns
 
 
-class SimpleRoleShareGPTPromptTokenizingStrategy(ShareGPTPromptTokenizingStrategy):
+class SimpleRoleShareGPTPromptTokenizingStrategy(
+    SimpleShareGPTPromptTokenizingStrategy
+):
     """
     basic sharegpt strategy to grab conversations from the sample row, but uses role instead of from
     """
@@ -209,3 +191,16 @@ class GlaiveShareGPTPromptTokenizingStrategy(SimpleShareGPTPromptTokenizingStrat
         conversation = merge_consecutive_messages(conversation)
 
         return conversation
+
+
+load = build_loader(SimpleShareGPTPromptTokenizingStrategy, ShareGPTPrompterV2)
+load_role = build_loader(SimpleRoleShareGPTPromptTokenizingStrategy, ShareGPTPrompterV2)
+load_ultrachat = build_loader(
+    UltrachatShareGPTPromptTokenizingStrategy, ShareGPTPrompterV2
+)
+load_guanaco = build_loader(GuanacoShareGPTPromptTokenizingStrategy, ShareGPTPrompterV2)
+load_glaive = build_loader(
+    GlaiveShareGPTPromptTokenizingStrategy,
+    ShareGPTPrompterV2,
+    default_conversation="chatml_glaive",
+)

src/axolotl/prompters.py

@@ -263,6 +263,7 @@ CONVERSATION_ROLE_FORMAT = {
     "chatml": "<|im_start|>{ROLE}",
     "zephyr": "<|{ROLE}|>",
     "vicuna_v1.1": "{ROLE}",
+    "llama3": "<|start_header_id|>{ROLE}<|end_header_id|>",
 }
 
 
@@ -348,7 +349,10 @@ class ShareGPTPrompter(Prompter):  # pylint: disable=too-few-public-methods
                 )
 
             if len(conv.messages) > 0 and ((role == conv.messages[-1][0])):
-                LOG.warning(f"{SHAREGPT_ASSERTION_FAILED_ROLE}: {sentence}")
+                if (
+                    role != "assistant"
+                ):  # back to back assistant calls may be okay for tool calls
+                    LOG.warning(f"{SHAREGPT_ASSERTION_FAILED_ROLE}: {sentence}")
 
             conv.append_message(role, sentence["value"])
 

src/axolotl/train.py

@@ -212,6 +212,10 @@ def train(
         if cfg.flash_optimum and BetterTransformer:
             model = BetterTransformer.reverse(model)
 
+        if cfg.rl and cfg.adapter and not cfg.rl_adapter_ref_model:
+            trainer.model.save_pretrained(
+                cfg.output_dir, safe_serialization=safe_serialization
+            )
         model.save_pretrained(cfg.output_dir, safe_serialization=safe_serialization)
 
     if not cfg.hub_model_id:

src/axolotl/utils/callbacks/__init__.py

@@ -778,6 +778,17 @@ class SaveAxolotlConfigtoWandBCallback(TrainerCallback):
 class SaveModelOnTrainEndCallback(TrainerCallback):
     """Callback to save model on train end"""
 
+    def on_step_end(  # pylint: disable=unused-argument
+        self,
+        args: TrainingArguments,
+        state: TrainerState,
+        control: TrainerControl,
+        **kwargs,
+    ):
+        # Save
+        if state.global_step >= state.max_steps:
+            control.should_save = True
+
     def on_train_end(  # pylint: disable=unused-argument
         self, args, state, control, **kwargs
     ):

src/axolotl/utils/chat_templates.py

@@ -24,6 +24,7 @@ def chat_templates(user_choice: str):
         "chatml": "{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% for message in messages %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
         "gemma": "{{ bos_token }}{% if messages[0]['role'] == 'system' %}{{ raise_exception('System role not supported') }}{% endif %}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if (message['role'] == 'assistant') %}{% set role = 'model' %}{% else %}{% set role = message['role'] %}{% endif %}{{ '<start_of_turn>' + role + '\n' + message['content'] | trim + '<end_of_turn>\n' }}{% endfor %}{% if add_generation_prompt %}{{'<start_of_turn>model\n'}}{% endif %}",
         "cohere": "{{ bos_token }}{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% elif false == true %}{% set loop_messages = messages %}{% set system_message = 'You are Command-R, a brilliant, sophisticated, AI-assistant trained to assist human users by providing thorough responses. You are trained by Cohere.' %}{% else %}{% set loop_messages = messages %}{% set system_message = false %}{% endif %}{% if system_message != false %}{{ '<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>' + system_message + '<|END_OF_TURN_TOKEN|>' }}{% endif %}{% for message in loop_messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '<|START_OF_TURN_TOKEN|><|USER_TOKEN|>' + content.strip() + '<|END_OF_TURN_TOKEN|>' }}{% elif message['role'] == 'assistant' %}{{ '<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>'  + content.strip() + '<|END_OF_TURN_TOKEN|>' }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>' }}{% endif %}",
+        "llama3": "{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% else %}{{ eos_token }}{% endif %}",
     }
 
     if user_choice in templates:

src/axolotl/utils/collators.py

@@ -229,9 +229,8 @@ class PretrainingBatchSamplerDataCollatorForSeq2Seq(DataCollatorForSeq2Seq):
             if feature == "attention_mask":
                 if self.multipack_attn:
                     arrays = [
-                        (i + 1) * np.array(item[feature])
+                        (i + 1) * np.array(item)
                         for i, item in enumerate(features[feature])
-                        if feature in item
                     ]
                 else:
                     arrays = [(1) * np.array(item) for item in features[feature]]

src/axolotl/utils/config/models/input/v0_4_1/__init__.py

@@ -133,7 +133,6 @@ class RLType(str, Enum):
     ipo = "ipo"  # pylint: disable=invalid-name
     kto_pair = "kto_pair"  # pylint: disable=invalid-name
     orpo = "orpo"  # pylint: disable=invalid-name
-    sppo_hard = "sppo_hard"  # pylint: disable=invalid-name
 
 
 class ChatTemplate(str, Enum):
@@ -144,6 +143,7 @@ class ChatTemplate(str, Enum):
     inst = "inst"  # pylint: disable=invalid-name
     gemma = "gemma"  # pylint: disable=invalid-name
     cohere = "cohere"  # pylint: disable=invalid-name
+    llama3 = "llama3"  # pylint: disable=invalid-name
 
 
 class LoftQConfig(BaseModel):
@@ -517,6 +517,9 @@ class AxolotlInputConfig(
 
     sequence_len: int = Field(default=512)
     min_sample_len: Optional[int] = None
+    max_prompt_len: int = Field(
+        default=512, metadata={"help": "maximum prompt length for RL training"}
+    )
     sample_packing: Optional[bool] = None
     eval_sample_packing: Optional[bool] = None
     pad_to_sequence_len: Optional[bool] = None
@@ -558,6 +561,8 @@ class AxolotlInputConfig(
     torch_compile: Optional[bool] = None
     torch_compile_backend: Optional[str] = None
 
+    custom_trainer_cls: Optional[str] = None
+
     max_steps: Optional[int] = None
     warmup_steps: Optional[int] = None
     warmup_ratio: Optional[float] = None
@@ -575,7 +580,6 @@ class AxolotlInputConfig(
     neftune_noise_alpha: Optional[float] = None
 
     orpo_alpha: Optional[float] = None
-    dpo_beta: Optional[float] = None
 
     max_memory: Optional[
         Dict[Union[int, Literal["cpu", "disk"]], Union[int, str]]

src/axolotl/utils/models.py

@@ -1,4 +1,5 @@
 """Module for models and model loading"""
+
 # pylint: disable=too-many-lines
 
 import logging
@@ -504,6 +505,9 @@ def load_model(
         bnb_config = {
             "load_in_8bit": True,
         }
+        # Exclude mamba blocks from int8 quantization for jamba
+        if cfg.model_config_type == "jamba":
+            bnb_config["llm_int8_skip_modules"] = ["mamba"]
         model_kwargs["quantization_config"] = BitsAndBytesConfig(
             **bnb_config,
         )
@@ -789,11 +793,7 @@ def load_model(
     if not reference_model or cfg.lora_model_dir:
         # if we're not loading the reference model, then we're loading the model for training
         # then the dpo trainer doesn't want the peft model loaded over it, it just wants the lora/peft config
-        if (
+        if cfg.adapter and cfg.rl in ["dpo", "ipo", "kto_pair"] and not cfg.merge_lora:
-            cfg.adapter
-            and cfg.rl in ["dpo", "ipo", "kto_pair", "sppo_hard"]
-            and not cfg.merge_lora
-        ):
             _, lora_config = load_lora(model, cfg, inference=False, config_only=True)
         else:
             model, lora_config = load_adapter(model, cfg, cfg.adapter)

src/axolotl/utils/trainer.py

@@ -438,7 +438,7 @@ def prepare_optim_env(cfg):
 
 
 def setup_trainer(cfg, train_dataset, eval_dataset, model, tokenizer, total_num_steps):
-    if cfg.rl in ["dpo", "ipo", "kto_pair", "orpo", "sppo_hard"]:
+    if cfg.rl in ["dpo", "ipo", "kto_pair", "orpo"]:
         trainer_builder = HFRLTrainerBuilder(cfg, model[0], tokenizer)
         trainer_builder.model_ref = model[1]
         trainer_builder.peft_config = model[2]

tests/prompt_strategies/test_sharegpt.py

@@ -12,10 +12,12 @@ from axolotl.prompt_strategies.sharegpt import (
     GlaiveShareGPTPromptTokenizingStrategy,
     SimpleShareGPTPromptTokenizingStrategy,
     register_chatml_template,
+    register_llama3_template,
 )
 from axolotl.prompters import ShareGPTPrompterV2
 
 register_chatml_template()
+register_llama3_template()
 
 
 @pytest.fixture(name="sharegpt_dataset")
@@ -115,7 +117,53 @@ def fixture_tokenizer():
     return tokenizer
 
 
-class TestSharegpt:
+@pytest.fixture(name="llama3_tokenizer")
+def fixture_llama3_tokenizer():
+    tokenizer = AutoTokenizer.from_pretrained("NousResearch/Meta-Llama-3-8B")
+    tokenizer.eos_token = "<|eot_id|>"
+
+    return tokenizer
+
+
+class TestSharegptLlama3:
+    """Test class for ShareGPT style datasets with llama-3 prompts"""
+
+    def test_tokenization(self, sharegpt_dataset, llama3_tokenizer):
+        strategy = SimpleShareGPTPromptTokenizingStrategy(
+            ShareGPTPrompterV2(
+                conversation="llama3",
+                role_key_model=None,
+                role_key_human=None,
+            ),
+            llama3_tokenizer,
+            False,  # train_on_inputs
+            2048,  # sequence_len
+        )
+
+        dataset_wrapper = TokenizedPromptDataset(
+            strategy, sharegpt_dataset, process_count=1
+        )
+
+        input_ids = dataset_wrapper[0]["input_ids"]
+
+        # fmt: off
+        assert input_ids == [
+            128000,  # bos
+            128006, 9125, 128007,  # system header
+            271, 31724, 128009,  # sys prompt, eot
+            128006, 882, 128007,  # user header
+            271, 15339, 128009,  # user prompt eot
+            128006, 78191, 128007,  # assistant header
+            271, 15339, 128009,   # assistant response eot
+            128006, 882, 128007,
+            271, 19045, 29474, 128009,
+            128006, 78191, 128007,
+            271, 19045, 29474, 128009,
+        ]
+        # fmt: on
+
+
+class TestSharegptChatML:
     """
     Test class for sharegpt prompter
     """