Training in progress, step 200

axolotl_config.yaml

@@ -1,5 +1,5 @@
 base_model: echarlaix/tiny-random-PhiForCausalLM
-batch_size: 32
+batch_size: 128
 bf16: true
 chat_template: tokenizer_default_fallback_alpaca
 datasets:
@@ -13,29 +13,35 @@ datasets:
     no_input_format: '{instruction}'
     system_format: '{system}'
     system_prompt: ''
-eval_steps: 20
+device_map: auto
+eval_sample_packing: false
+eval_steps: 200
 flash_attention: true
-gpu_memory_limit: 80GiB
+gradient_checkpointing: true
 group_by_length: true
 hub_model_id: SystemAdmin123/tiny-random-PhiForCausalLM
 hub_strategy: checkpoint
 learning_rate: 0.0002
 logging_steps: 10
 lr_scheduler: cosine
-max_steps: 2500
-micro_batch_size: 2
+max_steps: 10000
+micro_batch_size: 32
 model_type: AutoModelForCausalLM
 num_epochs: 100
 optimizer: adamw_bnb_8bit
-output_dir: /root/.sn56/axolotl/outputs/tiny-random-PhiForCausalLM
+output_dir: /root/.sn56/axolotl/tmp/tiny-random-PhiForCausalLM
 pad_to_sequence_len: true
 resize_token_embeddings_to_32x: false
-save_steps: 40
+sample_packing: true
+save_steps: 200
 save_total_limit: 1
 sequence_len: 2048
 special_tokens:
   pad_token: <|endoftext|>
 tokenizer_type: GPTNeoXTokenizerFast
+torch_dtype: bf16
+training_args_kwargs:
+  hub_private_repo: true
 trust_remote_code: true
 val_set_size: 0.1
 wandb_entity: ''

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:72b63ad3d536f40d7a1d427e171327ecf384565a6e7dce4b10b64c8015cacdbf
+oid sha256:b07bcd34427d8aa6f79f7f89126979a8c5aba7953e9406d3a5d723724a768f7d
 size 163534

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1d5ab278664d188b1e606142da773f714a65d45c121528cb5bcab2e692995c84
-size 6776
+oid sha256:a3e144aeaa0bf3508ddc68ca31b3eb1d164a15c9881c0609a3f6018245190804
+size 6904