End of training

README.md

@@ -1,13 +1,14 @@
 ---
 library_name: peft
-base_model: jingyeom/seal3.1.6n_7b
+license: llama3
+base_model: unsloth/llama-3-8b-Instruct
 tags:
 - axolotl
 - generated_from_trainer
 datasets:
-- Paladiso/dataset_2379ea32-7a7b-48aa-9667-b04a8ab9542e
+- Paladiso/dataset_695de20c-0af8-4b07-94bc-5ccdfcc25776
 model-index:
-- name: b0047e68-f838-43dc-b949-048a410f2378
+- name: a1ea3ebd-561e-45da-86b1-ff6386e13625
   results: []
 ---
 
@@ -20,17 +21,18 @@ should probably proofread and complete it, then remove this comment. -->
 axolotl version: `0.6.0`
 ```yaml
 adapter: lora
-base_model: jingyeom/seal3.1.6n_7b
+base_model: unsloth/llama-3-8b-Instruct
 bf16: auto
 chat_template: llama3
 dataset_prepared_path: /workspace/axolotl/data/prepared
 datasets:
 - ds_type: json
   format: custom
-  path: Paladiso/dataset_2379ea32-7a7b-48aa-9667-b04a8ab9542e
+  path: Paladiso/dataset_695de20c-0af8-4b07-94bc-5ccdfcc25776
   type:
-    field_instruction: question
-    field_output: answer
+    field_input: parent_id
+    field_instruction: role
+    field_output: text
     system_format: '{system}'
     system_prompt: ''
 debug: null
@@ -46,7 +48,7 @@ fsdp_config: null
 gradient_accumulation_steps: 4
 gradient_checkpointing: false
 group_by_length: false
-hub_model_id: Paladiso/b0047e68-f838-43dc-b949-048a410f2378
+hub_model_id: Paladiso/a1ea3ebd-561e-45da-86b1-ff6386e13625
 hub_private_repo: true
 hub_repo: null
 hub_strategy: checkpoint
@@ -86,10 +88,10 @@ use_accelerate: true
 val_set_size: 0.05
 wandb_entity: null
 wandb_mode: online
-wandb_name: 2379ea32-7a7b-48aa-9667-b04a8ab9542e
+wandb_name: 695de20c-0af8-4b07-94bc-5ccdfcc25776
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
-wandb_runid: 2379ea32-7a7b-48aa-9667-b04a8ab9542e
+wandb_runid: 695de20c-0af8-4b07-94bc-5ccdfcc25776
 warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
@@ -98,11 +100,11 @@ xformers_attention: null
 
 </details><br>
 
-# b0047e68-f838-43dc-b949-048a410f2378
+# a1ea3ebd-561e-45da-86b1-ff6386e13625
 
-This model is a fine-tuned version of [jingyeom/seal3.1.6n_7b](https://huggingface.co/jingyeom/seal3.1.6n_7b) on the Paladiso/dataset_2379ea32-7a7b-48aa-9667-b04a8ab9542e dataset.
+This model is a fine-tuned version of [unsloth/llama-3-8b-Instruct](https://huggingface.co/unsloth/llama-3-8b-Instruct) on the Paladiso/dataset_695de20c-0af8-4b07-94bc-5ccdfcc25776 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 3.3618
+- Loss: 2.0698
 
 ## Model description
 
@@ -136,9 +138,9 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 4.1771        | 0.0002 | 3    | 4.4973          |
-| 4.321         | 0.0004 | 6    | 4.1982          |
-| 3.7891        | 0.0006 | 9    | 3.3618          |
+| 2.4536        | 0.0011 | 3    | 2.3868          |
+| 2.1539        | 0.0021 | 6    | 2.3161          |
+| 1.9608        | 0.0032 | 9    | 2.0698          |
 
 
 ### Framework versions