make pad_to_sequence_len default to the same value as sample_packing (#2941) [skip ci]

* make pad_to_sequence_len default to the same value as sample_packing

* remove duplicate validation

* fix test

* update description meta

Co-authored-by: NanoCode012 <nano@axolotl.ai>

---------

Co-authored-by: NanoCode012 <nano@axolotl.ai>

examples/llama-3/3b-qat-fsdp2.yaml

@@ -22,7 +22,7 @@ datasets:
 output_dir: ./outputs/qat_out/
 
 sample_packing: true
-pad_to_sequence_len: true
+
 sequence_len: 512
 
 flex_attention: true

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

@@ -26,7 +26,7 @@ output_dir: ./outputs/out
 
 sequence_len: 4096
 sample_packing: true
-pad_to_sequence_len: true
+
 
 wandb_project:
 wandb_entity:

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

@@ -11,7 +11,7 @@ output_dir: ./outputs/out
 
 sequence_len: 8192
 sample_packing: true
-pad_to_sequence_len: true
+
 
 wandb_project:
 wandb_entity:

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

@@ -37,7 +37,7 @@ output_dir: ./outputs/lora-out
 
 sequence_len: 4096
 sample_packing: false
-pad_to_sequence_len: true
+
 
 adapter: lora
 lora_model_dir:

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

@@ -28,7 +28,7 @@ output_dir: ./outputs/lora-out
 
 sequence_len: 4096
 sample_packing: false
-pad_to_sequence_len: true
+
 
 adapter: lora
 lora_model_dir:

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

@@ -49,7 +49,7 @@ output_dir: ./outputs/lora-out
 
 sequence_len: 4096
 sample_packing: false
-pad_to_sequence_len: true
+
 
 adapter: lora
 lora_model_dir:

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

@@ -22,7 +22,7 @@ dataset_exact_deduplication: true
 sequence_len: 4096
 sample_packing: true
 eval_sample_packing: false
-pad_to_sequence_len: true
+
 
 adapter: lora
 lora_model_dir:

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

@@ -14,7 +14,7 @@ lora_model_dir:
 
 sequence_len: 2048
 sample_packing: true
-pad_to_sequence_len: true
+
 
 lora_r: 16
 lora_alpha: 32

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

@@ -15,7 +15,7 @@ lora_model_dir:
 sequence_len: 2048
 sample_packing: true
 eval_sample_packing: true
-pad_to_sequence_len: true
+
 
 lora_r: 16
 lora_alpha: 32

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

@@ -24,7 +24,7 @@ sample_packing: true
 sample_packing_sequentially: true
 curriculum_sampling: true
 eval_sample_packing: false
-pad_to_sequence_len: true
+
 
 adapter: lora
 lora_model_dir:

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

@@ -15,7 +15,7 @@ lora_model_dir:
 sequence_len: 2048
 sample_packing: true
 eval_sample_packing: true
-pad_to_sequence_len: true
+
 
 lora_r: 16
 lora_alpha: 32

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

@@ -18,7 +18,7 @@ output_dir: ./outputs/lora-out
 sequence_len: 4096
 sample_packing: true
 eval_sample_packing: false
-pad_to_sequence_len: true
+
 
 adapter: lora
 lora_model_dir:

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

@@ -18,7 +18,7 @@ lora_model_dir:
 sequence_len: 2048
 sample_packing: true
 eval_sample_packing: true
-pad_to_sequence_len: true
+
 
 lora_r: 32
 lora_alpha: 16

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

@@ -18,7 +18,7 @@ adapter: qlora
 
 sequence_len: 2048
 sample_packing: true
-pad_to_sequence_len: true
+
 
 lora_r: 16
 lora_alpha: 16

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

@@ -20,7 +20,7 @@ lora_model_dir:
 
 sequence_len: 512
 sample_packing: false
-pad_to_sequence_len: true
+
 
 lora_r: 8
 lora_alpha: 16

examples/llama-3/qlora.yml

@@ -20,7 +20,7 @@ lora_model_dir:
 
 sequence_len: 4096
 sample_packing: true
-pad_to_sequence_len: true
+
 
 lora_r: 32
 lora_alpha: 16

examples/llama-3/sparse-finetuning.yaml

@@ -16,7 +16,7 @@ output_dir: ./outputs/out
 
 sequence_len: 4096
 sample_packing: true
-pad_to_sequence_len: true
+
 eval_sample_packing: false
 
 wandb_project: