sqlai/sqlm

README.md

@@ -16,8 +16,6 @@ should probably proofread and complete it, then remove this comment. -->
 # sqlm
 
 This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) on the None dataset.
-It achieves the following results on the evaluation set:
-- Loss: 0.3241
 
 ## Model description
 
@@ -43,22 +41,10 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: constant
 - lr_scheduler_warmup_steps: 0.03
-- training_steps: 200
+- training_steps: 10
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss |
-|:-------------:|:-----:|:----:|:---------------:|
-| 0.5521        | 0.03  | 20   | 0.4444          |
-| 0.3942        | 0.05  | 40   | 0.3773          |
-| 0.3777        | 0.08  | 60   | 0.3571          |
-| 0.3452        | 0.11  | 80   | 0.3451          |
-| 0.3219        | 0.13  | 100  | 0.3431          |
-| 0.3339        | 0.16  | 120  | 0.3366          |
-| 0.3147        | 0.19  | 140  | 0.3334          |
-| 0.307         | 0.21  | 160  | 0.3341          |
-| 0.3283        | 0.24  | 180  | 0.3259          |
-| 0.3174        | 0.27  | 200  | 0.3241          |
 
 
 ### Framework versions

adapter_config.json

@@ -20,8 +20,8 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
-    "q_proj"
+    "q_proj",
+    "v_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:739125dd3903120a3d11a3ab51361a116a6035f8a7ec01147c26fcf7196b8b73
+oid sha256:5a6a096b32bd7f9b896c64ea0c801f4f004316fbcbffa2f2e438db5939a22925
 size 109069176

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9218cdac3e7a572c69e362fac2567a0149f1b3bbb61170b8ecc5f472a2ecb731
+oid sha256:98ac353eca3d0fdc1ab5fe79a338474eb151d63dffaea28ab8f10178f4b75cb9
 size 4856