enh: enable CodeLlama (#48)

predibase · Nov 21, 2023 · 5b1bb93 · 5b1bb93
1 parent 188834f
commit 5b1bb93
Show file tree

Hide file tree

Showing 2 changed files with 12 additions and 21 deletions.
diff --git a/server/lorax_server/models/flash_llama.py b/server/lorax_server/models/flash_llama.py
@@ -4,7 +4,7 @@
 
 from loguru import logger
 from opentelemetry import trace
-from transformers.models.llama import LlamaTokenizer, LlamaTokenizerFast
+from transformers import AutoTokenizer
 from tqdm import tqdm
 from typing import Dict, Optional
 
@@ -45,22 +45,13 @@ def __init__(
         else:
             raise NotImplementedError("FlashLlama is only available on GPU")
 
-        try:
-            tokenizer = LlamaTokenizer.from_pretrained(
-                model_id,
-                revision=revision,
-                padding_side="left",
-                truncation_side="left",
-                trust_remote_code=trust_remote_code,
-            )
-        except Exception:
-            tokenizer = LlamaTokenizerFast.from_pretrained(
-                model_id,
-                revision=revision,
-                padding_side="left",
-                truncation_side="left",
-                trust_remote_code=trust_remote_code,
-            )
+        tokenizer = AutoTokenizer.from_pretrained(
+            model_id,
+            revision=revision,
+            padding_side="left",
+            truncation_side="left",
+            trust_remote_code=trust_remote_code,
+        )
 
         config = LlamaConfig.from_pretrained(
             model_id, revision=revision, trust_remote_code=trust_remote_code

diff --git a/server/requirements.txt b/server/requirements.txt
@@ -1,4 +1,4 @@
-accelerate==0.19.0 ; python_version >= "3.9" and python_version < "4.0"
+accelerate==0.24.1 ; python_version >= "3.9" and python_version < "4.0"
 aiohttp==3.8.5 ; python_version >= "3.9" and python_version < "4.0"
 aiosignal==1.3.1 ; python_version >= "3.9" and python_version < "4.0"
 async-timeout==4.0.2 ; python_version >= "3.9" and python_version < "4.0"
@@ -23,7 +23,7 @@ grpcio-reflection==1.56.0 ; python_version >= "3.9" and python_version < "4.0"
 grpcio-status==1.56.0 ; python_version >= "3.9" and python_version < "4.0"
 grpcio==1.56.0 ; python_version >= "3.9" and python_version < "4.0"
 hf-transfer==0.1.3 ; python_version >= "3.9" and python_version < "4.0"
-huggingface-hub==0.14.1 ; python_version >= "3.9" and python_version < "4.0"
+huggingface-hub==0.19.4 ; python_version >= "3.9" and python_version < "4.0"
 idna==3.4 ; python_version >= "3.9" and python_version < "4.0"
 jinja2==3.1.2 ; python_version >= "3.9" and python_version < "4.0"
 loguru==0.6.0 ; python_version >= "3.9" and python_version < "4.0"
@@ -59,10 +59,10 @@ setuptools==68.0.0 ; python_version >= "3.9" and python_version < "4.0"
 six==1.16.0 ; python_version >= "3.9" and python_version < "4.0"
 sympy==1.12 ; python_version >= "3.9" and python_version < "4.0"
 texttable==1.6.7 ; python_version >= "3.9" and python_version < "4.0"
-tokenizers==0.13.3 ; python_version >= "3.9" and python_version < "4.0"
+tokenizers==0.15.0 ; python_version >= "3.9" and python_version < "4.0"
 torch==2.0.1 ; python_version >= "3.9" and python_version < "4.0"
 tqdm==4.65.0 ; python_version >= "3.9" and python_version < "4.0"
-transformers==4.29.2 ; python_version >= "3.9" and python_version < "4.0"
+transformers==4.35.2 ; python_version >= "3.9" and python_version < "4.0"
 typer==0.6.1 ; python_version >= "3.9" and python_version < "4.0"
 typing-extensions==4.7.1 ; python_version >= "3.9" and python_version < "4.0"
 tzdata==2023.3 ; python_version >= "3.9" and python_version < "4.0"