retrained models to fix bug in original training

vanbujm · Sep 27, 2024 · 3cb2bc3 · 3cb2bc3
1 parent 4d9917b
commit 3cb2bc3
Show file tree

Hide file tree

Showing 16 changed files with 31 additions and 53 deletions.
diff --git a/src/trian/train_cai.py b/src/trian/train_cai.py
@@ -141,4 +141,8 @@ def formatting_func(example):
 
 trainer.train()
 
-trainer.save_model("./ultrachat_baseline")
+trainer.save_model("./ultrachat_cai")
+
+trainer.push_to_hub("vanbujm/ultrachat_cai")
+tokenizer.push_to_hub("vanbujm/ultrachat_cai")
+
diff --git a/ultrachat_baseline/README.md b/ultrachat_baseline/README.md
@@ -1,6 +1,7 @@
 ---
 library_name: transformers
-tags: []
+tags:
+- unsloth
 ---
 
 # Model Card for Model ID

diff --git a/ultrachat_baseline/adapter_config.json b/ultrachat_baseline/adapter_config.json
@@ -20,13 +20,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
-    "down_proj",
-    "up_proj",
     "o_proj",
     "k_proj",
+    "up_proj",
     "q_proj",
-    "gate_proj"
+    "gate_proj",
+    "down_proj",
+    "v_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

diff --git a/ultrachat_baseline/adapter_model.safetensors b/ultrachat_baseline/adapter_model.safetensors
diff --git a/ultrachat_baseline/special_tokens_map.json b/ultrachat_baseline/special_tokens_map.json
@@ -13,13 +13,7 @@
     "rstrip": false,
     "single_word": false
   },
-  "pad_token": {
-    "content": "</s>",
-    "lstrip": false,
-    "normalized": false,
-    "rstrip": false,
-    "single_word": false
-  },
+  "pad_token": "<unk>",
   "unk_token": {
     "content": "<unk>",
     "lstrip": false,

diff --git a/ultrachat_baseline/tokenizer.json b/ultrachat_baseline/tokenizer.json
@@ -1,11 +1,6 @@
 {
   "version": "1.0",
-  "truncation": {
-    "direction": "Right",
-    "max_length": 4096,
-    "strategy": "LongestFirst",
-    "stride": 0
-  },
+  "truncation": null,
   "padding": null,
   "added_tokens": [
     {

diff --git a/ultrachat_baseline/tokenizer.model b/ultrachat_baseline/tokenizer.model
diff --git a/ultrachat_baseline/tokenizer_config.json b/ultrachat_baseline/tokenizer_config.json
@@ -34,15 +34,12 @@
   "clean_up_tokenization_spaces": false,
   "eos_token": "</s>",
   "legacy": false,
-  "max_length": 4096,
-  "model_max_length": 1000000000000000019884624838656,
-  "pad_token": "</s>",
+  "model_max_length": 32768,
+  "pad_token": "<unk>",
+  "padding_side": "right",
   "sp_model_kwargs": {},
   "spaces_between_special_tokens": false,
-  "stride": 0,
   "tokenizer_class": "LlamaTokenizer",
-  "truncation_side": "right",
-  "truncation_strategy": "longest_first",
   "unk_token": "<unk>",
   "use_default_system_prompt": false
 }
diff --git a/ultrachat_baseline/training_args.bin b/ultrachat_baseline/training_args.bin
diff --git a/ultrachat_cai/README.md b/ultrachat_cai/README.md
@@ -1,6 +1,7 @@
 ---
 library_name: transformers
-tags: []
+tags:
+- unsloth
 ---
 
 # Model Card for Model ID

diff --git a/ultrachat_cai/adapter_config.json b/ultrachat_cai/adapter_config.json
@@ -20,13 +20,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
+    "v_proj",
     "down_proj",
-    "gate_proj",
-    "q_proj",
     "k_proj",
+    "q_proj",
+    "o_proj",
     "up_proj",
-    "v_proj",
-    "o_proj"
+    "gate_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

diff --git a/ultrachat_cai/adapter_model.safetensors b/ultrachat_cai/adapter_model.safetensors
diff --git a/ultrachat_cai/special_tokens_map.json b/ultrachat_cai/special_tokens_map.json
@@ -13,13 +13,7 @@
     "rstrip": false,
     "single_word": false
   },
-  "pad_token": {
-    "content": "</s>",
-    "lstrip": false,
-    "normalized": false,
-    "rstrip": false,
-    "single_word": false
-  },
+  "pad_token": "<unk>",
   "unk_token": {
     "content": "<unk>",
     "lstrip": false,

diff --git a/ultrachat_cai/tokenizer.json b/ultrachat_cai/tokenizer.json
@@ -1,11 +1,6 @@
 {
   "version": "1.0",
-  "truncation": {
-    "direction": "Right",
-    "max_length": 4096,
-    "strategy": "LongestFirst",
-    "stride": 0
-  },
+  "truncation": null,
   "padding": null,
   "added_tokens": [
     {

diff --git a/ultrachat_cai/tokenizer_config.json b/ultrachat_cai/tokenizer_config.json
@@ -34,15 +34,12 @@
   "clean_up_tokenization_spaces": false,
   "eos_token": "</s>",
   "legacy": false,
-  "max_length": 4096,
-  "model_max_length": 1000000000000000019884624838656,
-  "pad_token": "</s>",
+  "model_max_length": 32768,
+  "pad_token": "<unk>",
+  "padding_side": "right",
   "sp_model_kwargs": {},
   "spaces_between_special_tokens": false,
-  "stride": 0,
   "tokenizer_class": "LlamaTokenizer",
-  "truncation_side": "right",
-  "truncation_strategy": "longest_first",
   "unk_token": "<unk>",
   "use_default_system_prompt": false
 }
diff --git a/ultrachat_cai/training_args.bin b/ultrachat_cai/training_args.bin