Training in progress, step 100

axolotl_config.yaml

@@ -1,5 +1,5 @@
 base_model: fxmarty/tiny-llama-fast-tokenizer
-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-llama-fast-tokenizer
 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-llama-fast-tokenizer
+output_dir: /root/.sn56/axolotl/tmp/tiny-llama-fast-tokenizer
 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: </s>
 tokenizer_type: LlamaTokenizerFast
+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 sha256:0d9f1e60d34df70b7758959048fe7f2701ecbcbb16b25defdf06c629d7d60c03
+oid sha256:25a257c5b2e34eda80e3d84f1a8cc4247ba163d63b057609915a20d5c03487fe
 size 2066752

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5fd5f69a596f4120903b9be7edad5a6b39384d6ec2e365296723a2d6d4a993e7
-size 6712
+oid sha256:aaa4e96900978d3ffc6b4d538660eb0cf2fc137143c61a37b1d477787b8740ea
+size 6904