End of training

README.md

@@ -1,14 +1,14 @@
 ---
 library_name: peft
-license: llama3
-base_model: NousResearch/Hermes-3-Llama-3.1-8B
+license: mit
+base_model: numind/NuExtract-v1.5
 tags:
 - axolotl
 - generated_from_trainer
 datasets:
-- Paladiso/dataset_01e3a775-b4d1-4728-83d2-a8ae99e1f4fe
+- Paladiso/dataset_c99871fd-1142-4a68-9841-64b293e07c60
 model-index:
-- name: 5cafc7e9-7d7b-4f3f-a5a3-7358b1931d00
+- name: 6a2139b5-d27e-477f-8b4a-9be9f04e75b5
   results: []
 ---
 
@@ -21,17 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
 axolotl version: `0.6.0`
 ```yaml
 adapter: lora
-base_model: NousResearch/Hermes-3-Llama-3.1-8B
+base_model: numind/NuExtract-v1.5
 bf16: auto
 chat_template: llama3
 dataset_prepared_path: /workspace/axolotl/data/prepared
 datasets:
 - ds_type: json
   format: custom
-  path: Paladiso/dataset_01e3a775-b4d1-4728-83d2-a8ae99e1f4fe
+  path: Paladiso/dataset_c99871fd-1142-4a68-9841-64b293e07c60
   type:
-    field_instruction: darija
-    field_output: english
+    field_instruction: instruction
+    field_output: output
     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: Paladiso/5cafc7e9-7d7b-4f3f-a5a3-7358b1931d00
+hub_model_id: Paladiso/6a2139b5-d27e-477f-8b4a-9be9f04e75b5
 hub_private_repo: true
 hub_repo: null
 hub_strategy: checkpoint
@@ -87,10 +87,10 @@ use_accelerate: true
 val_set_size: 0.05
 wandb_entity: null
 wandb_mode: online
-wandb_name: 01e3a775-b4d1-4728-83d2-a8ae99e1f4fe
+wandb_name: c99871fd-1142-4a68-9841-64b293e07c60
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
-wandb_runid: 01e3a775-b4d1-4728-83d2-a8ae99e1f4fe
+wandb_runid: c99871fd-1142-4a68-9841-64b293e07c60
 warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
@@ -99,11 +99,11 @@ xformers_attention: null
 
 </details><br>
 
-# 5cafc7e9-7d7b-4f3f-a5a3-7358b1931d00
+# 6a2139b5-d27e-477f-8b4a-9be9f04e75b5
 
-This model is a fine-tuned version of [NousResearch/Hermes-3-Llama-3.1-8B](https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-8B) on the Paladiso/dataset_01e3a775-b4d1-4728-83d2-a8ae99e1f4fe dataset.
+This model is a fine-tuned version of [numind/NuExtract-v1.5](https://huggingface.co/numind/NuExtract-v1.5) on the Paladiso/dataset_c99871fd-1142-4a68-9841-64b293e07c60 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 2.2292
+- Loss: 1.9610
 
 ## Model description
 
@@ -137,9 +137,9 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 2.5678        | 0.0090 | 3    | 2.7981          |
-| 2.5156        | 0.0180 | 6    | 2.6932          |
-| 1.9508        | 0.0271 | 9    | 2.2292          |
+| 8.2516        | 0.0021 | 3    | 2.1313          |
+| 7.4297        | 0.0041 | 6    | 2.0895          |
+| 7.3104        | 0.0062 | 9    | 1.9610          |
 
 
 ### Framework versions