End of training

README.md

@@ -26,9 +26,9 @@ strict: false
 
 chat_template: chatml
 datasets:
-  # - path: lordspline/scidata
-  #   type: sharegpt
-  #   conversation: chatml
+  - path: lordspline/scidata
+    type: sharegpt
+    conversation: chatml
   - path: lordspline/wizard_v2_196k_unfiltered
     type: sharegpt
     conversation: chatml
@@ -80,7 +80,7 @@ warmup_steps: 10
 evals_per_epoch: 1
 eval_table_size:
 eval_max_new_tokens: 128
-saves_per_epoch: 100
+saves_per_epoch: 300
 debug:
 
 # deepspeed: deepspeed_configs/zero3_bf16.json
@@ -101,7 +101,7 @@ tokens:
 
 This model is a fine-tuned version of [lordspline/mergestein](https://huggingface.co/lordspline/mergestein) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.0348
+- Loss: 1.2069
 
 ## Model description
 
@@ -133,7 +133,7 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step  | Validation Loss |
 |:-------------:|:-----:|:-----:|:---------------:|
-| 1.1202        | 1.0   | 25552 | 1.0348          |
+| 1.4476        | 1.0   | 48435 | 1.2069          |
 
 
 ### Framework versions

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fc89906f67ddbeb76efbb37239aa493a9b881a2577994f05671aea509adc0188
+oid sha256:33b1343643141d6472066aaebabd297f82b0f61ecd310ecb18d4d7c616dc8cb0
 size 1589947346