Merge pull request #746 from riedgar-ms/riedgar-ms/llama-cpp-ub-remove

[Test] llama-cpp-python update

.github/workflows/unit_tests.yml

@@ -37,7 +37,7 @@ jobs:
           if [ -f requirements.txt ]; then pip install -r requirements.txt; fi
       - name: Install model-specific dependencies
         run: |
-          pip install "llama-cpp-python<0.2.58"
+          pip install "llama-cpp-python!=0.2.58"
       - name: Run tests (except server)
         shell: bash
         run: |

.github/workflows/unit_tests_gpu.yml

@@ -50,7 +50,7 @@ jobs:
       - name: GPU pip installs
         run: |
           pip install accelerate
-          CMAKE_ARGS="-DLLAMA_CUBLAS=on" pip install "llama-cpp-python<0.2.58"
+          CMAKE_ARGS="-DLLAMA_CUBLAS=on" pip install "llama-cpp-python!=0.2.58"
       - name: Check GPU available
         run: |
           python -c "import torch; assert torch.cuda.is_available()"

guidance/models/llama_cpp/_llama_cpp.py

@@ -118,7 +118,7 @@ def __init__(self, model, compute_log_probs, **kwargs):
                 )
 
             with normalize_notebook_stdout_stderr():
-                self.model_obj = llama_cpp.Llama(model_path=model, **kwargs)
+                self.model_obj = llama_cpp.Llama(model_path=model, logits_all=True, **kwargs)
         elif isinstance(model, llama_cpp.Llama):
             self.model = model.__class__.__name__
             self.model_obj = model