End of training

README.md

@@ -35,10 +35,10 @@ This model was trained with SFT.
 
 ### Framework versions
 
-- TRL: 0.14.0
-- Transformers: 4.48.2
+- TRL: 0.15.2
+- Transformers: 4.49.0
 - Pytorch: 2.6.0
-- Datasets: 3.2.0
+- Datasets: 3.3.2
 - Tokenizers: 0.21.0
 
 ## Citations

adapter_config.json

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

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:65b5dc25a48f1723ea790f54d5ba3ecbbffdf7273ef9a138abf21de61c9987d7
+oid sha256:0c1dedb72088a73a439cfe44734f2115f5ab27e2c24ab48682efc1de990d8c59
 size 194563400

runs/Feb15_22-30-26_ea6e1e24e8ea/events.out.tfevents.1739658631.ea6e1e24e8ea.190592.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:babeeae9699d20246d3c41fa4f992b1be263d35708176b84f6e96640b966ad01
+size 6520

runs/Feb18_04-06-24_ea6e1e24e8ea/events.out.tfevents.1739851588.ea6e1e24e8ea.682413.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f0ab2be7dea5eec2ebba129cb5c2ae6c65d94b3fb1f3aa9ca7485144c5e1372
+size 4184

runs/Feb18_04-08-29_ea6e1e24e8ea/events.out.tfevents.1739851713.ea6e1e24e8ea.683084.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44a41353e30c2a88e7428d0cac08838d3bc788f57874593588158d4f00876897
+size 4184

runs/Feb18_04-10-40_ea6e1e24e8ea/events.out.tfevents.1739851843.ea6e1e24e8ea.683763.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53c7b4700679a64d678a9748e61d15c7065422b509e7d606d81f30a669e3fedd
+size 74577

runs/Mar13_03-21-28_1c541e6483bb/events.out.tfevents.1741836092.1c541e6483bb.6671.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55c209a95d598c26ebd3f3b35604ebfdad9c68a1b399bd372cc8c690d277d7c9
+size 74609

runs/Mar13_03-44-30_1c541e6483bb/events.out.tfevents.1741837472.1c541e6483bb.11177.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a338969601f123f5e8a33f2d62d1b28078c5ac7b080723c33c74ea1b6e0f6b2
+size 33491

tokenizer_config.json

@@ -2061,7 +2061,7 @@
   ],
   "model_max_length": 131072,
   "pad_token": "<|finetune_right_pad_id|>",
-  "padding_side": "left",
-  "tokenizer_class": "PreTrainedTokenizerFast",
+  "padding_side": "right",
+  "tokenizer_class": "PreTrainedTokenizer",
   "unk_token": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1328e5a15bc595f9802adc522b62ad33513a8ebc15081fd3148d104d1e91bc21
-size 5624
+oid sha256:bb18cffe9997604e0e9fdfdadbf0bc70f88b93ae383f6fba5f372cf14b4f9e48
+size 5688