EdBerg/SmolLM2-1.7B-Instruct_Baha_1

README.md

@@ -27,7 +27,7 @@ print(output["generated_text"])
 
 ## Training procedure
 
-[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/harpermia882/huggingface/runs/c1yo0njb)
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/harpermia882/huggingface/runs/9z7s33ce)
 
 This model was trained with SFT.
 

adapter_config.json

@@ -24,13 +24,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "down_proj",
-    "q_proj",
-    "v_proj",
-    "o_proj",
+    "gate_proj",
     "up_proj",
     "k_proj",
-    "gate_proj"
+    "v_proj",
+    "q_proj",
+    "o_proj",
+    "down_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:a223001d84c3a8d683af2393c54df900cef306a6c9a965b8135ef7ba685014db
+oid sha256:75cb2c847ea9a42324fe42149ea06c5b7d321001f588e312128998b4af2b5235
 size 144748392

runs/Feb13_00-04-04_c0908c98e147/events.out.tfevents.1739405046.c0908c98e147.193.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0711829d906f5c168a6e2a37a75213febf196c2dde4f9b1bf6fe6e2b78f5faca
+size 6130

runs/Feb13_00-39-53_c0908c98e147/events.out.tfevents.1739407194.c0908c98e147.193.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57c754930a5a76cc18ac8a111d2c75265c7906efb23cf73d5091a520f93e70e5
+size 6128

runs/Feb13_00-41-42_c0908c98e147/events.out.tfevents.1739407304.c0908c98e147.193.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b50a5e1ef49ab781d32a8eb0c0fcb3467d15b55d92577de48a4a5d9aa6194ad
+size 6128

runs/Feb13_00-42-30_c0908c98e147/events.out.tfevents.1739407351.c0908c98e147.193.3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55184a38c8f3c4a49d1d3c1edfb7b40c1db8235d7cee6f0814f4adf21f10f218
+size 6129

runs/Feb13_00-43-43_c0908c98e147/events.out.tfevents.1739407423.c0908c98e147.193.4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1378e9c364a8bf3d5c5f33d8aaf3d0e76bb607e6bce0df583a567dba5b479be4
+size 4184

runs/Feb13_00-46-41_c0908c98e147/events.out.tfevents.1739407602.c0908c98e147.12468.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1e87bacccc01bb73303e1b6b781f7d0e8ddc8ca7f3f7b324abe0be5aa2d3335
+size 27177

special_tokens_map.json

@@ -17,7 +17,13 @@
     "rstrip": false,
     "single_word": false
   },
-  "pad_token": "<|im_end|>",
+  "pad_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
   "unk_token": {
     "content": "<|endoftext|>",
     "lstrip": false,

tokenizer.json

@@ -2,7 +2,7 @@
   "version": "1.0",
   "truncation": {
     "direction": "Right",
-    "max_length": 256,
+    "max_length": 1024,
     "strategy": "LongestFirst",
     "stride": 0
   },

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5616a592153ca222ac29515ef5cb32bd321b1b8793b11e869c0743e20630c3f3
+oid sha256:12cd2b0fd4c5f2ec0aa665ea6d6d239d5b27e06098f0bd8f491cd87138d52d93
 size 5624