Ignore

.gitignore

@@ -1,3 +1,4 @@
+/hf_model/
 /Model/
 /Tokenizer/
 *.pth

config.json

@@ -1,11 +0,0 @@
-{
-  "_attn_implementation_autoset": true,
-  "block_size": 1024,
-  "dropout": 0.2,
-  "model_type": "gpt_custom",
-  "n_embd": 768,
-  "n_head": 12,
-  "n_layer": 12,
-  "transformers_version": "4.46.2",
-  "vocab_size": 20000
-}

modeling_gpt_custom.py

@@ -1,33 +0,0 @@
-from transformers import PreTrainedModel, PretrainedConfig, AutoModel, AutoConfig
-import torch
-from Naive_gpt.model import GPTLanguageModel 
-
-class GPTConfig(PretrainedConfig):
-    model_type = "gpt_custom"
-
-    def __init__(self, vocab_size=20000, **kwargs):
-        super().__init__(**kwargs)
-        self.vocab_size = vocab_size
-        self.block_size = 1024
-        self.n_embd = 768
-        self.n_head = 12
-        self.n_layer = 12
-        self.dropout = 0.2
-
-class GPTModelHF(PreTrainedModel):
-    config_class = GPTConfig
-
-    def __init__(self, config):
-        super().__init__(config)
-        self.model = GPTLanguageModel(config.vocab_size, config)
-        self.config = config
-
-    def forward(self, x):
-        return self.model(x)
-
-    def generate(self, idx, max_new_tokens):
-        return self.model.generate(idx, max_new_tokens)
-
-# Register the model (this will run when the file is imported)
-AutoConfig.register("gpt_custom", GPTConfig)
-AutoModel.register(GPTConfig, GPTModelHF)

pytorch_model.bin

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:98621e405ea217da4b3906ae0ebe10862d9225d7ca9479ed714a0576e407b31e
-size 404940190

tokenizer_config.json

@@ -1,4 +0,0 @@
-{
-    "type": "llama",
-    "vocab_size": 20000
-}