End of training

README.md

@@ -14,7 +14,7 @@ datasets:
 - coastral/korean-writing-style-instruct
 - devngho/korean-instruction-mix
 model-index:
-- name: ko-tiny-exp
+- name: tiny-ko-sft
   results: []
 ---
 
@@ -28,6 +28,11 @@ axolotl version: `0.10.0.dev0`
 ```yaml
 base_model: minpeter/pretrained-tiny-ko
 
+hub_model_id: minpeter/tiny-ko-sft
+output_dir: ./outputs/tiny-ko-sft
+wandb_project: "axolotl"
+wandb_entity: "kasfiekfs-e"
+
 chat_template: chatml
 datasets:
   - path: lemon-mint/Korean-FineTome-100k
@@ -102,11 +107,6 @@ datasets:
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.05
 
-hub_model_id: minpeter/ko-tiny-exp
-output_dir: ./ouputs/ko-tiny-exp
-wandb_project: "axolotl"
-wandb_entity: "kasfiekfs-e"
-
 save_steps: 200
 warmup_steps: 20
 eval_steps: 200
@@ -153,11 +153,11 @@ weight_decay: 0.0
 
 </details><br>
 
-# ko-tiny-exp
+# tiny-ko-sft
 
 This model is a fine-tuned version of [minpeter/pretrained-tiny-ko](https://huggingface.co/minpeter/pretrained-tiny-ko) on the lemon-mint/Korean-FineTome-100k, the lemon-mint/smol-koreantalk, the heegyu/open-korean-instructions-v20231020, the FreedomIntelligence/evol-instruct-korean, the FreedomIntelligence/alpaca-gpt4-korean, the FreedomIntelligence/sharegpt-korean, the coastral/korean-writing-style-instruct and the devngho/korean-instruction-mix datasets.
 It achieves the following results on the evaluation set:
-- Loss: 1.4634
+- Loss: 1.4286
 
 ## Model description
 
@@ -194,21 +194,21 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 2.696         | 0.0010 | 1    | 2.7432          |
-| 1.7677        | 0.2019 | 200  | 1.7528          |
-| 1.6696        | 0.4037 | 400  | 1.6833          |
-| 1.5866        | 0.6056 | 600  | 1.6401          |
-| 1.6249        | 0.8075 | 800  | 1.5957          |
-| 1.3578        | 1.0091 | 1000 | 1.5704          |
-| 1.4469        | 1.2110 | 1200 | 1.5514          |
-| 1.3969        | 1.4128 | 1400 | 1.5220          |
-| 1.3549        | 1.6147 | 1600 | 1.4939          |
-| 1.3107        | 1.8166 | 1800 | 1.4695          |
-| 1.2462        | 2.0182 | 2000 | 1.4751          |
-| 1.2001        | 2.2200 | 2200 | 1.4692          |
-| 1.0911        | 2.4219 | 2400 | 1.4661          |
-| 1.1547        | 2.6238 | 2600 | 1.4636          |
-| 1.1943        | 2.8256 | 2800 | 1.4634          |
+| 2.9956        | 0.0010 | 1    | 3.0182          |
+| 1.7162        | 0.2019 | 200  | 1.7023          |
+| 1.6186        | 0.4037 | 400  | 1.6351          |
+| 1.5474        | 0.6056 | 600  | 1.5951          |
+| 1.5822        | 0.8075 | 800  | 1.5540          |
+| 1.3144        | 1.0091 | 1000 | 1.5333          |
+| 1.403         | 1.2110 | 1200 | 1.5128          |
+| 1.3558        | 1.4128 | 1400 | 1.4832          |
+| 1.3165        | 1.6147 | 1600 | 1.4541          |
+| 1.2704        | 1.8166 | 1800 | 1.4305          |
+| 1.1913        | 2.0182 | 2000 | 1.4424          |
+| 1.1488        | 2.2200 | 2200 | 1.4346          |
+| 1.0417        | 2.4219 | 2400 | 1.4311          |
+| 1.1104        | 2.6238 | 2600 | 1.4288          |
+| 1.1446        | 2.8256 | 2800 | 1.4286          |
 
 
 ### Framework versions