Update model configuration and fix tokenizer files; remove outdated model binary

logs/poetry_generation_20241117.log

@@ -2,3 +2,19 @@
 2024-11-17 00:08:50,303 - main - INFO - Model and tokenizer loaded successfully
 2024-11-17 00:13:06,341 - main - INFO - Initializing model on device: cpu
 2024-11-17 00:13:07,660 - main - INFO - Model and tokenizer loaded successfully
+2024-11-17 16:33:11,148 - main - INFO - Initializing model on device: cpu
+2024-11-17 16:33:13,017 - main - ERROR - Error initializing model: Error(s) in loading state_dict for GPT2LMHeadModel:
+	size mismatch for transformer.wpe.weight: copying a param with shape torch.Size([400, 384]) from checkpoint, the shape in current model is torch.Size([128, 384]).
+2024-11-17 16:33:13,017 - main - ERROR - Detailed traceback:
+Traceback (most recent call last):
+  File "E:\Self Work\My Projects\Poetica HuggingFace Server\poetica\main.py", line 137, in initialize
+    await self._load_and_optimize_model()
+  File "E:\Self Work\My Projects\Poetica HuggingFace Server\poetica\main.py", line 185, in _load_and_optimize_model
+    self.model.load_state_dict(state_dict, strict=False)
+  File "e:\Self Work\My Projects\Poetica HuggingFace Server\.venv\Lib\site-packages\torch\nn\modules\module.py", line 2189, in load_state_dict
+    raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
+RuntimeError: Error(s) in loading state_dict for GPT2LMHeadModel:
+	size mismatch for transformer.wpe.weight: copying a param with shape torch.Size([400, 384]) from checkpoint, the shape in current model is torch.Size([128, 384]).
+2024-11-17 16:33:13,020 - main - ERROR - Failed to initialize model manager
+2024-11-17 16:33:41,008 - main - INFO - Initializing model on device: cpu
+2024-11-17 16:33:43,152 - main - INFO - Model and tokenizer loaded successfully

main.py

@@ -23,8 +23,8 @@ BATCH_SIZE = 4
 CACHE_SIZE = 1024
 
 MODEL_CONFIG = GPT2Config(
-    n_positions=128,
-    n_ctx=128,
+    n_positions=400,
+    n_ctx=400,
     n_embd=384,
     n_layer=6,
     n_head=6,

models/poeticagpt.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f77da9534fcf01b36f4780cd24ebe46e4d7f8740a1b17b66d5173d8694d6a62e
-size 139310252
+oid sha256:3ba4e77d7a7b5186188172eb2559210305b4e565459e84b8ddadd26a63a0ebbf
+size 139728044

models/pytorch_model.bin

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f77da9534fcf01b36f4780cd24ebe46e4d7f8740a1b17b66d5173d8694d6a62e
-size 139310252

models/special_tokens_map.json

@@ -21,4 +21,4 @@
     "rstrip": false,
     "single_word": false
   }
-}
+}

models/tokenizer_config.json

@@ -19,4 +19,4 @@
   "pad_token": "<|endoftext|>",
   "tokenizer_class": "GPT2Tokenizer",
   "unk_token": "<|endoftext|>"
-}
+}

models/vocab.json

@@ -50256,4 +50256,4 @@
   "Ń": 255,
   "Ń·": 48953,
   "ŃĶ": 18433
-}
+}