End of training

README.md

@@ -1,14 +1,14 @@
 ---
 library_name: peft
-license: apache-2.0
-base_model: NousResearch/Yarn-Mistral-7b-128k
+license: llama3
+base_model: NousResearch/Hermes-3-Llama-3.1-8B
 tags:
 - axolotl
 - generated_from_trainer
 datasets:
-- Aivesa/dataset_0e0354d0-8f7f-4af8-a88c-0b79cf1c515d
+- Aivesa/dataset_01e3a775-b4d1-4728-83d2-a8ae99e1f4fe
 model-index:
-- name: 3ac905c6-b56d-4d35-810a-855512cee15e
+- name: 4cfb1824-c33e-4bbb-ae66-a0fc7539cc92
   results: []
 ---
 
@@ -21,18 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
 axolotl version: `0.6.0`
 ```yaml
 adapter: lora
-base_model: NousResearch/Yarn-Mistral-7b-128k
+base_model: NousResearch/Hermes-3-Llama-3.1-8B
 bf16: auto
 chat_template: llama3
 dataset_prepared_path: /workspace/axolotl/data/prepared
 datasets:
 - ds_type: json
   format: custom
-  path: Aivesa/dataset_0e0354d0-8f7f-4af8-a88c-0b79cf1c515d
+  path: Aivesa/dataset_01e3a775-b4d1-4728-83d2-a8ae99e1f4fe
   type:
-    field_input: Text
-    field_instruction: Task
-    field_output: Generation
+    field_instruction: darija
+    field_output: english
     system_format: '{system}'
     system_prompt: ''
 debug: null
@@ -48,7 +47,7 @@ fsdp_config: null
 gradient_accumulation_steps: 4
 gradient_checkpointing: false
 group_by_length: false
-hub_model_id: Aivesa/3ac905c6-b56d-4d35-810a-855512cee15e
+hub_model_id: Aivesa/4cfb1824-c33e-4bbb-ae66-a0fc7539cc92
 hub_private_repo: true
 hub_repo: null
 hub_strategy: checkpoint
@@ -79,8 +78,6 @@ sample_packing: false
 save_safetensors: true
 saves_per_epoch: 4
 sequence_len: 512
-special_tokens:
-  pad_token: </s>
 strict: false
 tf32: false
 tokenizer_type: AutoTokenizer
@@ -90,10 +87,10 @@ use_accelerate: true
 val_set_size: 0.05
 wandb_entity: null
 wandb_mode: online
-wandb_name: 0e0354d0-8f7f-4af8-a88c-0b79cf1c515d
+wandb_name: 01e3a775-b4d1-4728-83d2-a8ae99e1f4fe
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
-wandb_runid: 0e0354d0-8f7f-4af8-a88c-0b79cf1c515d
+wandb_runid: 01e3a775-b4d1-4728-83d2-a8ae99e1f4fe
 warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
@@ -102,11 +99,11 @@ xformers_attention: null
 
 </details><br>
 
-# 3ac905c6-b56d-4d35-810a-855512cee15e
+# 4cfb1824-c33e-4bbb-ae66-a0fc7539cc92
 
-This model is a fine-tuned version of [NousResearch/Yarn-Mistral-7b-128k](https://huggingface.co/NousResearch/Yarn-Mistral-7b-128k) on the Aivesa/dataset_0e0354d0-8f7f-4af8-a88c-0b79cf1c515d dataset.
+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 Aivesa/dataset_01e3a775-b4d1-4728-83d2-a8ae99e1f4fe dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.6436
+- Loss: 2.2331
 
 ## Model description
 
@@ -140,9 +137,9 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 8.6059        | 0.0020 | 3    | 2.1353          |
-| 8.1233        | 0.0040 | 6    | 1.9151          |
-| 6.2762        | 0.0059 | 9    | 1.6436          |
+| 2.5688        | 0.0090 | 3    | 2.7949          |
+| 2.5156        | 0.0180 | 6    | 2.6941          |
+| 1.9482        | 0.0271 | 9    | 2.2331          |
 
 
 ### Framework versions