End of training

README.md

@@ -1,13 +1,14 @@
 ---
 library_name: peft
-base_model: NousResearch/CodeLlama-13b-hf
+license: llama3
+base_model: DeepMount00/Llama-3-8b-Ita
 tags:
 - axolotl
 - generated_from_trainer
 datasets:
-- Aivesa/dataset_4bea4578-524c-4430-8071-bbcf24df00eb
+- Aivesa/dataset_d14c0776-b5b3-443c-87b9-01747ff6b6aa
 model-index:
-- name: e2777369-66dd-42d7-90e2-868e01499551
+- name: 6de393f4-ef2f-45d5-9093-ef6e0404bb59
   results: []
 ---
 
@@ -20,18 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
 axolotl version: `0.6.0`
 ```yaml
 adapter: lora
-base_model: NousResearch/CodeLlama-13b-hf
+base_model: DeepMount00/Llama-3-8b-Ita
 bf16: auto
 chat_template: llama3
 dataset_prepared_path: /workspace/axolotl/data/prepared
 datasets:
 - ds_type: json
   format: custom
-  path: Aivesa/dataset_4bea4578-524c-4430-8071-bbcf24df00eb
+  path: Aivesa/dataset_d14c0776-b5b3-443c-87b9-01747ff6b6aa
   type:
-    field_input: text
-    field_instruction: intent
-    field_output: slots
+    field_instruction: "\u985E\u5225"
+    field_output: text
     system_format: '{system}'
     system_prompt: ''
 debug: null
@@ -47,7 +47,7 @@ fsdp_config: null
 gradient_accumulation_steps: 4
 gradient_checkpointing: false
 group_by_length: false
-hub_model_id: Aivesa/e2777369-66dd-42d7-90e2-868e01499551
+hub_model_id: Aivesa/6de393f4-ef2f-45d5-9093-ef6e0404bb59
 hub_private_repo: true
 hub_repo: null
 hub_strategy: checkpoint
@@ -79,7 +79,7 @@ save_safetensors: true
 saves_per_epoch: 4
 sequence_len: 512
 special_tokens:
-  pad_token: </s>
+  pad_token: <|eot_id|>
 strict: false
 tf32: false
 tokenizer_type: AutoTokenizer
@@ -89,10 +89,10 @@ use_accelerate: true
 val_set_size: 0.05
 wandb_entity: null
 wandb_mode: online
-wandb_name: 4bea4578-524c-4430-8071-bbcf24df00eb
+wandb_name: d14c0776-b5b3-443c-87b9-01747ff6b6aa
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
-wandb_runid: 4bea4578-524c-4430-8071-bbcf24df00eb
+wandb_runid: d14c0776-b5b3-443c-87b9-01747ff6b6aa
 warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
@@ -101,11 +101,11 @@ xformers_attention: null
 
 </details><br>
 
-# e2777369-66dd-42d7-90e2-868e01499551
+# 6de393f4-ef2f-45d5-9093-ef6e0404bb59
 
-This model is a fine-tuned version of [NousResearch/CodeLlama-13b-hf](https://huggingface.co/NousResearch/CodeLlama-13b-hf) on the Aivesa/dataset_4bea4578-524c-4430-8071-bbcf24df00eb dataset.
+This model is a fine-tuned version of [DeepMount00/Llama-3-8b-Ita](https://huggingface.co/DeepMount00/Llama-3-8b-Ita) on the Aivesa/dataset_d14c0776-b5b3-443c-87b9-01747ff6b6aa dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.2549
+- Loss: 2.8180
 
 ## Model description
 
@@ -139,9 +139,9 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 13.5248       | 0.0045 | 3    | 2.8330          |
-| 12.1843       | 0.0089 | 6    | 2.5730          |
-| 6.9279        | 0.0134 | 9    | 1.2549          |
+| 3.3319        | 0.0059 | 3    | 3.3605          |
+| 3.1276        | 0.0119 | 6    | 3.1618          |
+| 2.8718        | 0.0178 | 9    | 2.8180          |
 
 
 ### Framework versions