ctranslate2

config.json

@@ -3,8 +3,10 @@
     "DeciCoderForCausalLM"
   ],
   "auto_map": {
-    "AutoConfig": "configuration_decicoder.DeciCoderConfig",
-    "AutoModelForCausalLM": "modeling_decicoder.DeciCoderForCausalLM"
+    "AutoConfig": "configuration_llama.LlamaConfig",
+    "AutoModel": "modeling_llama.LlamaModel",
+    "AutoModelForCausalLM": "modeling_llama.LlamaForCausalLM",
+    "AutoModelForSequenceClassification": "modeling_llama.LlamaForSequenceClassification"
   },
   "bos_token_id": 0,
   "eos_token_id": 0,
@@ -25,4 +27,4 @@
   "transformers_version": "4.31.0.dev0",
   "use_cache": true,
   "vocab_size": 49152
-}
+}

configuration_decicoder.py → configuration_llama.py

@@ -13,7 +13,7 @@ logger = logging.get_logger(__name__)
 LLAMA_PRETRAINED_CONFIG_ARCHIVE_MAP = {}
 
 
-class DeciCoderConfig(LlamaConfig):
+class LlamaConfig(LlamaConfig):
     r"""
    This is the configuration class to store the configuration of a [`LlamaModel`]. It is used to instantiate an LLaMA
    model according to the specified arguments, defining the model architecture. Instantiating a configuration with the

ctranslate2/config.json

@@ -0,0 +1,6 @@
+{
+  "bos_token": "<|endoftext|>",
+  "eos_token": "<|endoftext|>",
+  "layer_norm_epsilon": 1e-05,
+  "unk_token": "<|endoftext|>"
+}

ctranslate2/model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d15385c0845adf975befc076f7eafd3a12059e0a89f3d18cd24b6610a5f66edb
+size 2227193677

tabby.json

@@ -0,0 +1,3 @@
+{
+    "auto_model": "AutoModelForCausalLM"
+}