OumaymaeELBIACH/Results_deepseek

README.md

@@ -36,7 +36,7 @@ This model was trained with SFT.
 ### Framework versions
 
 - TRL: 0.15.2
-- Transformers: 4.50.0.dev0
+- Transformers: 4.48.3
 - Pytorch: 2.6.0
 - Datasets: 3.3.2
 - Tokenizers: 0.21.0

adapter_config.json

@@ -23,13 +23,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "up_proj",
-    "v_proj",
-    "k_proj",
     "q_proj",
-    "o_proj",
+    "k_proj",
+    "gate_proj",
+    "v_proj",
     "down_proj",
-    "gate_proj"
+    "up_proj",
+    "o_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9a280212a87ecbe40215e680a0da0afb70c86edf51f3baeac3214bf21bf23be0
-size 167832688
+oid sha256:3ce11984e6f3f75a3f13bd963e7444794891b9194a31db67b51996d4cd9605b2
+size 335604696

runs/Mar03_10-34-48_e034797fa02e/events.out.tfevents.1740998218.e034797fa02e.2745.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb03d7bd4b5035b7a540bfd63d26def39ccf5f8af8aa1ff93143d17dd508b6c0
+size 9739

runs/Mar03_10-34-48_e034797fa02e/events.out.tfevents.1740998521.e034797fa02e.2745.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9bf50df259a90e79bda831304d2f7823d7684ab7d5e8ed566b4465ead9d6603
+size 416

tokenizer_config.json

@@ -2062,7 +2062,7 @@
   "pad_token": "<｜end▁of▁sentence｜>",
   "padding_side": "right",
   "sp_model_kwargs": {},
-  "tokenizer_class": "LlamaTokenizerFast",
+  "tokenizer_class": "LlamaTokenizer",
   "unk_token": null,
   "use_default_system_prompt": false
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:876d327f73d2ac58b19b4e7ae76e0927898e76e887e73267847bcbd155fecd17
-size 5688
+oid sha256:823132ccdae25cc2a7dd50a182ef4083a022dc7834e011e881a13f5e953984f3
+size 5624