Training in progress, step 50

axolotl_config.yaml

@@ -15,26 +15,26 @@ datasets:
     system_prompt: ''
 device_map: auto
 eval_sample_packing: false
-eval_steps: 200
+eval_steps: 50
 flash_attention: true
-gpu_memory_limit: 80GiB
+gradient_checkpointing: true
 group_by_length: true
 hub_model_id: SystemAdmin123/Phi-3.5-mini-instruct
 hub_strategy: checkpoint
 learning_rate: 0.0002
 logging_steps: 10
 lr_scheduler: cosine
-max_steps: 2500
+max_steps: 5000
 micro_batch_size: 4
 model_type: AutoModelForCausalLM
 num_epochs: 100
 optimizer: adamw_bnb_8bit
-output_dir: /root/.sn56/axolotl/outputs/Phi-3.5-mini-instruct
+output_dir: /root/.sn56/axolotl/tmp/Phi-3.5-mini-instruct
 pad_to_sequence_len: true
 resize_token_embeddings_to_32x: false
-sample_packing: false
-save_steps: 400
-save_total_limit: 1
+sample_packing: true
+save_steps: 50
+save_total_limit: 2
 sequence_len: 2048
 tokenizer_type: LlamaTokenizerFast
 torch_dtype: bf16

model-00001-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:571ac0ce3c6fb0674780b87152c4c23763691f3402fcfee2a7fae382cd311dc6
+oid sha256:826e66b930379b9b17d06e3e36abceae89de64258529cf378786e7914578f2d6
 size 4972163696

model-00002-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4a7f7e729b06122d8124d7ec438aa26d437fde3841f4d31cd4051d0537bd7093
+oid sha256:e2037f966379e933b0ee4ec2c31021e10d1bac8c66ea2bd6a5ed50ad79df2c79
 size 2669366920

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5d80a39228c54a221be1a2fcaa8735bc8f219e67dfabd2fc6349782d71ba757b
-size 7032
+oid sha256:e4320434255846d7d3d8916bcc913b473a3b0890bc89a39d373bf3ce9e989285
+size 7096