chore: refactor normalize_attn to use mapping and loop

* fsdp embeddings should be float32 per comment

* patch peft to not upcast everything

* add tabs back to code check

* fix import

* add configurable option and fix check

* add check for dtypes

* move embeddings test to patch dir

* fix test

* fix comment and logic

docs/config.qmd

@@ -32,6 +32,8 @@ tokenizer_legacy:
 resize_token_embeddings_to_32x:
 # Optional[bool] Whether to shrink the embeddings to len(tokenizer). By default, we won't shrink.
 shrink_embeddings:
+# Optional[bool] Don't upcast the embeddings to float32 when using PEFT. Useful for low-VRAM GPUs
+embeddings_skip_upcast:
 # Whether to load the model with randomly initialized weights. Useful for
 # pre-training a model from scratch or debugging purposes.
 random_init_weights:
@@ -73,11 +75,12 @@ load_in_8bit: true
 load_in_4bit:
 
 # Use CUDA bf16
-bf16: true # bool or 'full' for `bf16_full_eval`. require >=ampere
+bf16: true # bool or 'full' for `bf16_full_eval`, or 'auto' for automatic detection. require >=ampere
 # Use CUDA fp16
 fp16: true
 # Use CUDA tf32
 tf32: true # require >=ampere
+# Note: if bf16 is set to 'auto', and fp16 is set to true, we will prefer the explict fp16 setting
 
 # No AMP (automatic mixed precision)
 bfloat16: true # require >=ampere
@@ -184,8 +187,8 @@ datasets:
     # adding a system turn with empty content.
     drop_system_message:
 
-    # Optional[bool]. Whether to split the assistant turn based on a reasoning trace inside delimited tags
+    # Optional[bool]. (for Qwen3 template only) Whether to split the assistant content based on a reasoning trace inside delimited tags
-    # defaults to False
+    # See example at `docs/dataset-formats/conversation.qmd`
     split_thinking:
 
     # IMPORTANT: The following fields determine which parts of the conversation to train on.

docs/dataset-formats/conversation.qmd

@@ -196,6 +196,34 @@ datasets:
 It is not necessary to set both `message_field_training` and `message_field_training_detail` at once.
 :::
 
+8. (For Qwen3 template only) Enable reasoning split, where the reasoning is split from the content and passed as a separate field into the template.
+
+```yaml
+datasets:
+  - path: ...
+    type: chat_template
+    chat_template: qwen3
+    split_thinking: true
+```
+
+For example, a content can look like:
+
+```json
+{
+  "content": "<think>Some thinking outputs</think>Output after thinking."
+}
+```
+
+After split, it will look like:
+
+```json
+{
+  "reasoning_content": "Some thinking outputs",
+  "content": "Output after thinking..."
+}
+```
+
+
 ## sharegpt
 
 ::: {.callout-important}

examples/cerebras/btlm-ft.yml

@@ -59,9 +59,7 @@ gradient_checkpointing: false
 resume_from_checkpoint:
 logging_steps: 1
 
-flash_attention: true
+attention: flash
-sdp_attention:
-flash_optimum:
 
 gptq_groupsize:
 gptq_model_v1:

examples/cerebras/qlora.yml

@@ -39,8 +39,7 @@ tf32: true
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-xformers_attention: true
+attention: xformers
-flash_attention:
 gptq_groupsize:
 gptq_model_v1:
 warmup_steps: 10

examples/code-llama/13b/lora.yml

@@ -45,7 +45,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/code-llama/13b/qlora.yml

@@ -46,7 +46,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/code-llama/34b/lora.yml

@@ -45,7 +45,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/code-llama/34b/qlora.yml

@@ -46,7 +46,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/code-llama/7b/lora.yml

@@ -45,7 +45,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/code-llama/7b/qlora.yml

@@ -46,7 +46,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/cohere/command-r-7b-qlora.yml

@@ -49,7 +49,8 @@ tf32: true
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_ratio: 0.1
 evals_per_epoch:

examples/colab-notebooks/colab-axolotl-example.ipynb

@@ -112,9 +112,7 @@
     "early_stopping_patience:\n",
     "resume_from_checkpoint:\n",
     "logging_steps: 1\n",
-    "xformers_attention:\n",
+    "attention: sdpa\n",
-    "flash_attention: false\n",
-    "sdp_attention: true\n",
     "\n",
     "warmup_steps: 1\n",
     "max_steps: 25\n",

examples/dbrx/16bit-lora.yaml

@@ -52,7 +52,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch:

examples/dbrx/8bit-lora.yaml

@@ -55,7 +55,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch:

examples/dbrx/fft-ds-zero3.yaml

@@ -39,7 +39,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch:

examples/deepseek-v2/fft-fsdp-16b.yaml

@@ -35,7 +35,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 100
 evals_per_epoch: 2

examples/deepseek-v2/qlora-fsdp-2_5.yaml

@@ -59,7 +59,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 100
 evals_per_epoch: 2

examples/falcon/config-7b-lora.yml

@@ -43,8 +43,7 @@ tf32: true
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-xformers_attention: true
+attention: xformers
-flash_attention:
 gptq_groupsize:
 gptq_model_v1:
 warmup_steps: 40

examples/falcon/config-7b-qlora.yml

@@ -73,8 +73,7 @@ early_stopping_patience: 3
 resume_from_checkpoint:
 auto_resume_from_checkpoints: true
 logging_steps: 1
-xformers_attention: true
+attention: xformers
-flash_attention:
 gptq_groupsize:
 gptq_model_v1:
 warmup_steps: 10

examples/falcon/config-7b.yml

@@ -40,8 +40,7 @@ tf32: true
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-xformers_attention: true
+attention: xformers
-flash_attention:
 gptq_groupsize:
 gptq_model_v1:
 warmup_steps: 40

examples/gemma/qlora.yml

@@ -47,7 +47,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_ratio: 0.1
 evals_per_epoch: 4

examples/gemma2/qlora.yml

@@ -53,7 +53,8 @@ tf32: true
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_ratio: 0.1
 evals_per_epoch:

examples/gemma2/reward-model.yaml

@@ -43,7 +43,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_ratio: 0.1
 evals_per_epoch:

examples/gemma3/gemma-3-1b-qlora.yml

@@ -57,7 +57,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_ratio: 0.1
 evals_per_epoch:

examples/gemma3/gemma-3-4b-qlora.yml

@@ -51,8 +51,7 @@ gradient_checkpointing: true
 gradient_checkpointing_kwargs:
   use_reentrant: false
 logging_steps: 1
-flash_attention: true
+attention: flash
-eager_attention:
 
 warmup_ratio: 0.1
 evals_per_epoch: 1

examples/gemma3/gemma-3-4b-vision-qlora.yml

@@ -53,8 +53,7 @@ gradient_checkpointing: true
 gradient_checkpointing_kwargs:
   use_reentrant: false
 logging_steps: 1
-flash_attention: true
+attention: flash
-eager_attention:
 
 warmup_ratio: 0.1
 evals_per_epoch: 1

examples/gptj/qlora.yml

@@ -36,8 +36,7 @@ tf32: true
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-xformers_attention: true
+attention: xformers
-flash_attention:
 gptq_groupsize:
 gptq_model_v1:
 warmup_steps: 10

examples/jamba/qlora.yaml

@@ -47,7 +47,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch:

examples/jamba/qlora_deepspeed.yaml

@@ -46,7 +46,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch:

examples/jamba/qlora_fsdp_large.yaml

@@ -45,7 +45,8 @@ gradient_checkpointing: true
 gradient_checkpointing_kwargs:
   use_reentrant: true
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 1

examples/jeopardy-bot/config.yml

@@ -37,8 +37,7 @@ bf16: auto
 tf32: true
 resume_from_checkpoint:
 logging_steps: 5
-xformers_attention: true
+attention: xformers
-flash_attention:
 gptq_groupsize:
 gptq_model_v1:
 warmup_steps: 20

examples/llama-2/fft_optimized.yml

@@ -42,7 +42,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 flash_attn_cross_entropy: false
 flash_attn_rms_norm: true
 flash_attn_fuse_qkv: false

examples/llama-2/gptq-lora.yml

@@ -53,9 +53,7 @@ tf32: true
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention:
+attention: flash
-sdp_attention:
-flash_optimum:
 warmup_steps: 100
 evals_per_epoch: 4
 saves_per_epoch: 1

examples/llama-2/lisa.yml

@@ -46,7 +46,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 flash_attn_cross_entropy: false
 flash_attn_rms_norm: true
 flash_attn_fuse_qkv: false

examples/llama-2/loftq.yml

@@ -45,7 +45,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/llama-2/lora.yml

@@ -45,7 +45,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/llama-2/qlora-fsdp.yml

@@ -48,7 +48,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/llama-2/qlora.yml

@@ -46,7 +46,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/llama-2/relora.yml

@@ -48,7 +48,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/llama-3-vision/lora-11b.yaml

@@ -50,8 +50,7 @@ tf32: true
 
 gradient_checkpointing: true
 logging_steps: 1
-flash_attention: true
+attention: flash
-eager_attention:
 
 warmup_ratio: 0.1
 evals_per_epoch: 1

examples/llama-3/fft-8b-liger-fsdp.yaml

@@ -49,7 +49,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 100
 evals_per_epoch: 2

examples/llama-3/fft-8b.yaml

@@ -34,7 +34,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 100
 evals_per_epoch: 2

examples/llama-3/instruct-dpo-lora-8b.yml

@@ -61,7 +61,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/llama-3/instruct-lora-8b.yml

@@ -56,7 +56,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/llama-3/lora-1b-deduplicate-dpo.yml

@@ -77,7 +77,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/llama-3/lora-1b-deduplicate-sft.yml

@@ -53,7 +53,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/llama-3/lora-1b-kernels.yml

@@ -54,7 +54,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 loss_watchdog_threshold: 5.0
 loss_watchdog_patience: 3

examples/llama-3/lora-1b-ray.yml

@@ -48,7 +48,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 loss_watchdog_threshold: 5.0
 loss_watchdog_patience: 3

examples/llama-3/lora-1b-sample-packing-sequentially.yml

@@ -55,7 +55,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/llama-3/lora-1b.yml

@@ -48,7 +48,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 loss_watchdog_threshold: 5.0
 loss_watchdog_patience: 3

examples/llama-3/lora-8b.yml

@@ -49,7 +49,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/llama-3/qlora-1b-kto.yaml

@@ -53,7 +53,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 20
 evals_per_epoch: 4

examples/llama-3/qlora-1b.yml

@@ -51,7 +51,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 loss_watchdog_threshold: 5.0
 loss_watchdog_patience: 3

examples/llama-3/qlora-fsdp-405b.yaml

@@ -39,7 +39,8 @@ gradient_checkpointing: true
 gradient_checkpointing_kwargs:
   use_reentrant: true
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/llama-3/qlora-fsdp-70b.yaml

@@ -48,7 +48,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/llama-3/qlora.yml

@@ -46,7 +46,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/llava/lora-7b.yaml

@@ -46,8 +46,7 @@ tf32: true
 
 gradient_checkpointing: true
 logging_steps: 1
-flash_attention: true
+attention: flash
-eager_attention:
 
 warmup_ratio: 0.1
 evals_per_epoch: 1

examples/mamba/config.yml

@@ -39,7 +39,7 @@ tf32: true
 gradient_checkpointing: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention:
+attention: eager
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/mistral/bigstral-ds-zero3.yaml

@@ -42,7 +42,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 save_total_limit: 1
 save_steps:

examples/mistral/config.yml

@@ -36,7 +36,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/mistral/lora-mps.yml

@@ -53,8 +53,7 @@ tf32: true
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: false
+attention: sdpa
-sdp_attention: true
 
 loss_watchdog_threshold: 5.0
 loss_watchdog_patience: 3

examples/mistral/lora.yml

@@ -54,7 +54,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 loss_watchdog_threshold: 5.0
 loss_watchdog_patience: 3

examples/mistral/mistral-dpo-qlora.yml

@@ -71,7 +71,7 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: false
+attention: eager
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/mistral/mistral-qlora-fsdp.yml

@@ -51,7 +51,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 loss_watchdog_threshold: 5.0
 loss_watchdog_patience: 3

examples/mistral/mistral-qlora-orpo.yml

@@ -59,7 +59,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 loss_watchdog_threshold: 5.0
 loss_watchdog_patience: 3

examples/mistral/mistral-small-3.1-24B-lora.yml

@@ -48,9 +48,7 @@ tf32: true
 
 gradient_checkpointing: true
 logging_steps: 1
-flash_attention: false # PixtralVisionModel does not support Flash Attention 2.0 yet.
+attention: eager  # PixtralVisionModel does not support Flash Attention 2.0 yet.
-eager_attention:
-
 warmup_ratio: 0.1
 evals_per_epoch: 1
 saves_per_epoch: 1

examples/mistral/mixtral-8x22b-qlora-fsdp.yml

@@ -49,7 +49,8 @@ tf32: true
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 loss_watchdog_threshold: 5.0
 loss_watchdog_patience: 3

examples/mistral/mixtral-qlora-fsdp.yml

@@ -51,7 +51,8 @@ tf32: true
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 loss_watchdog_threshold: 5.0
 loss_watchdog_patience: 3

examples/mistral/mixtral.yml

@@ -69,7 +69,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 loss_watchdog_threshold: 5.0
 loss_watchdog_patience: 3

examples/mistral/mixtral_22.yml

@@ -40,7 +40,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 save_total_limit: 1
 save_steps:

examples/mistral/qlora.yml

@@ -54,7 +54,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 loss_watchdog_threshold: 5.0
 loss_watchdog_patience: 3

examples/mpt-7b/config.yml

@@ -39,7 +39,7 @@ bf16: auto
 tf32: true
 resume_from_checkpoint:
 logging_steps: 5
-flash_attention:
+attention: eager
 gptq_groupsize:
 gptq_model_v1:
 warmup_steps: 20

examples/openllama-3b/config.yml

@@ -39,7 +39,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 gptq_groupsize:
 gptq_model_v1:
 warmup_steps: 20

examples/openllama-3b/lora.yml

@@ -47,7 +47,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 gptq_groupsize:
 gptq_model_v1:
 warmup_steps: 20

examples/openllama-3b/qlora.yml

@@ -40,7 +40,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 gptq_groupsize:
 gptq_model_v1:
 warmup_steps: 20

examples/phi/phi-ft.yml

@@ -48,7 +48,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: True
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 100
 evals_per_epoch: 4

examples/phi/phi-qlora.yml

@@ -51,7 +51,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: True
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 100
 evals_per_epoch: 4

examples/phi/phi2-ft.yml

@@ -48,7 +48,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: True
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 100
 evals_per_epoch: 4

examples/phi/phi3-ft-fsdp.yml

@@ -49,7 +49,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 100
 evals_per_epoch: 4

examples/phi/phi3-ft.yml

@@ -44,7 +44,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: True
 early_stopping_patience: 3
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 eval_steps: 1000
 save_steps: 5000

examples/pixtral/lora-12b.yml

@@ -46,8 +46,7 @@ tf32: true
 
 gradient_checkpointing: true
 logging_steps: 1
-flash_attention: false # PixtralVisionModel does not support Flash Attention 2.0 yet
+attention: eager  # PixtralVisionModel does not support Flash Attention 2.0 yet
-eager_attention:
 
 warmup_ratio: 0.1
 evals_per_epoch: 1

examples/qwen/lora.yml

@@ -47,7 +47,7 @@ tf32: false
 gradient_checkpointing: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention:
+attention: eager
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/qwen/qlora.yml

@@ -47,7 +47,7 @@ tf32: false
 gradient_checkpointing: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention:
+attention: flash
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/qwen/qwen2-moe-lora.yaml

@@ -43,7 +43,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/qwen/qwen2-moe-qlora.yaml

@@ -46,7 +46,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/qwen2-vl/lora-7b.yaml

@@ -46,8 +46,7 @@ tf32: true
 
 gradient_checkpointing: true
 logging_steps: 1
-flash_attention: true
+attention: flash
-eager_attention:
 
 warmup_ratio: 0.1
 evals_per_epoch: 1

examples/qwen2/dpo.yaml

@@ -49,7 +49,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/qwen2/prm.yaml

@@ -47,7 +47,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_ratio: 0.1
 evals_per_epoch:

examples/qwen2/qlora-fsdp.yaml

@@ -47,7 +47,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/qwen2/reward-model.yaml

@@ -43,7 +43,8 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_ratio: 0.1
 evals_per_epoch:

examples/redpajama/config-3b.yml

@@ -40,7 +40,7 @@ bf16: auto
 tf32: true
 resume_from_checkpoint:
 logging_steps: 5
-flash_attention:
+attention: flash
 gptq_groupsize:
 gptq_model_v1:
 warmup_steps: 20

examples/replit-3b/config-lora.yml

@@ -38,7 +38,7 @@ tf32: true
 gradient_checkpointing:
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention:
+attention: eager
 gptq_groupsize:
 gptq_model_v1:
 warmup_steps: 20

examples/stablelm-2/1.6b/fft.yml

@@ -44,7 +44,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 flash_attn_cross_entropy: false
 flash_attn_rms_norm: true
 flash_attn_fuse_qkv: false

examples/stablelm-2/1.6b/lora.yml

@@ -47,7 +47,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 flash_attn_cross_entropy: false
 flash_attn_rms_norm: true
 

examples/starcoder2/qlora.yml

@@ -46,7 +46,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 20
 evals_per_epoch: 4

examples/tiny-llama/lora-mps.yml

@@ -47,7 +47,7 @@ tf32: true
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: false
+attention: eager
 
 warmup_steps: 10
 evals_per_epoch: 0

examples/tiny-llama/lora.yml

@@ -45,7 +45,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/tiny-llama/pretrain.yml

@@ -36,7 +36,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch:

examples/tiny-llama/qlora.yml

@@ -47,7 +47,8 @@ tf32: false
 gradient_checkpointing: true
 resume_from_checkpoint:
 logging_steps: 1
-flash_attention: true
+attention: flash
+
 
 warmup_steps: 10
 evals_per_epoch: 4

examples/xgen-7b/xgen-7b-8k-qlora.yml

@@ -71,8 +71,7 @@ early_stopping_patience: 3
 resume_from_checkpoint:
 auto_resume_from_checkpoints: true
 logging_steps: 1
-xformers_attention: true
+attention: xformers
-flash_attention:
 gptq_groupsize:
 gptq_model_v1:
 warmup_steps: 10