Merge pull request #1434 from Codium-ai/tr/ollama

Tr/ollama

docs/docs/usage-guide/changing_a_model.md

@@ -30,50 +30,36 @@ model="" # the OpenAI model you've deployed on Azure (e.g. gpt-4o)
 fallback_models=["..."]
 ```
 
-### Hugging Face
+### Ollama
 
-**Local**
-You can run Hugging Face models locally through either [VLLM](https://docs.litellm.ai/docs/providers/vllm) or [Ollama](https://docs.litellm.ai/docs/providers/ollama)
+You can run models locally through either [VLLM](https://docs.litellm.ai/docs/providers/vllm) or [Ollama](https://docs.litellm.ai/docs/providers/ollama)
 
-E.g. to use a new Hugging Face model locally via Ollama, set:
+E.g. to use a new model locally via Ollama, set in `.secrets.toml` or in a configuration file:
 ```
-[__init__.py]
-MAX_TOKENS = {
-    "model-name-on-ollama": <max_tokens>
-}
-e.g.
-MAX_TOKENS={
-    ...,
-    "ollama/llama2": 4096
-}
-
-
-[config] # in configuration.toml
-model = "ollama/llama2"
-fallback_models=["ollama/llama2"]
+[config]
+model = "ollama/qwen2.5-coder:32b"
+fallback_models=["ollama/qwen2.5-coder:32b"]
+custom_model_max_tokens=128000 # set the maximal input tokens for the model
+duplicate_examples=true # will duplicate the examples in the prompt, to help the model to output structured output
 
-[ollama] # in .secrets.toml
-api_base = ... # the base url for your Hugging Face inference endpoint
-# e.g. if running Ollama locally, you may use:
-api_base = "http://localhost:11434/"
+[ollama]
+api_base = "http://localhost:11434" # or whatever port you're running Ollama on
 ```
 
-### Inference Endpoints
+!!! note "Local models vs commercial models"
+    Qodo Merge is compatible with almost any AI model, but analyzing complex code repositories and pull requests requires a model specifically optimized for code analysis.
+    Commercial models such as GPT-4, Claude Sonnet, and Gemini have demonstrated robust capabilities in generating structured output for code analysis tasks with large input. In contrast, most open-source models currently available (as of January 2025) face challenges with these complex tasks.
+    Based on our testing, local open-source models are suitable for experimentation and learning purposes, but they are not suitable for production-level code analysis tasks.
+    Hence, for production workflows and real-world usage, we recommend using commercial models.
+
+### Hugging Face Inference Endpoints
 
 To use a new model with Hugging Face Inference Endpoints, for example, set:
 ```
-[__init__.py]
-MAX_TOKENS = {
-    "model-name-on-huggingface": <max_tokens>
-}
-e.g.
-MAX_TOKENS={
-    ...,
-    "meta-llama/Llama-2-7b-chat-hf": 4096
-}
 [config] # in configuration.toml
 model = "huggingface/meta-llama/Llama-2-7b-chat-hf"
 fallback_models=["huggingface/meta-llama/Llama-2-7b-chat-hf"]
+custom_model_max_tokens=... # set the maximal input tokens for the model
 
 [huggingface] # in .secrets.toml
 key = ... # your Hugging Face api key

pr_agent/config_loader.py

@@ -12,7 +12,6 @@
     envvar_prefix=False,
     merge_enabled=True,
     settings_files=[join(current_dir, f) for f in [
-        "settings/.secrets.toml",
         "settings/configuration.toml",
         "settings/ignore.toml",
         "settings/language_extensions.toml",
@@ -29,6 +28,7 @@
         "settings/pr_add_docs.toml",
         "settings/custom_labels.toml",
         "settings/pr_help_prompts.toml",
+        "settings/.secrets.toml",
         "settings_prod/.secrets.toml",
     ]]
 )

pr_agent/settings/configuration.toml

@@ -34,6 +34,7 @@ ai_disclaimer_title=""  # Pro feature, title for a collapsible disclaimer to AI
 ai_disclaimer=""  # Pro feature, full text for the AI disclaimer
 output_relevant_configurations=false
 large_patch_policy = "clip" # "clip", "skip"
+duplicate_prompt_examples = false
 # seed
 seed=-1 # set positive value to fix the seed (and ensure temperature=0)
 temperature=0.2

pr_agent/settings/pr_code_suggestions_prompts.toml

@@ -125,6 +125,30 @@ The PR Diff:
 {{ diff_no_line_numbers|trim }}
 ======
 
+{%- if duplicate_prompt_examples %}
+
+
+Example output:
+```yaml
+code_suggestions:
+- relevant_file: |
+    src/file1.py
+  language: |
+    python
+  suggestion_content: |
+    ...
+  existing_code: |
+    ...
+  improved_code: |
+    ...
+  one_sentence_summary: |
+    ...
+  label: |
+    ...
+```
+(replace '...' with actual content)
+{%- endif %}
+
 
 Response (should be a valid YAML, and nothing else):
 ```yaml

pr_agent/settings/pr_code_suggestions_reflect_prompts.toml

@@ -122,6 +122,25 @@ Below are {{ num_code_suggestions }} AI-generated code suggestions for enhancing
 ======
 
 
+{%- if duplicate_prompt_examples %}
+
+
+Example output:
+```yaml
+code_suggestions:
+- suggestion_summary: |
+    ...
+  relevant_file: "..."
+  relevant_lines_start: ...
+  relevant_lines_end: ...
+  suggestion_score: ...
+  why: |
+    ...
+- ...
+```
+(replace '...' with actual content)
+{%- endif %}
+
 Response (should be a valid YAML, and nothing else):
 ```yaml
 """

pr_agent/settings/pr_description_prompts.toml

@@ -130,6 +130,37 @@ The PR Git Diff:
 
 Note that lines in the diff body are prefixed with a symbol that represents the type of change: '-' for deletions, '+' for additions, and ' ' (a space) for unchanged lines.
 
+{%- if duplicate_prompt_examples %}
+
+
+Example output:
+```yaml
+type:
+- Bug fix
+- Refactoring
+- ...
+description: |
+  ...
+title: |
+  ...
+{%- if enable_semantic_files_types %}
+pr_files:
+- filename: |
+    ...
+{%- if include_file_summary_changes %}
+  changes_summary: |
+    ...
+{%- endif %}
+  changes_title: |
+    ...
+  label: |
+    label_key_1
+...
+{%- endif %}
+```
+(replace '...' with the actual values)
+{%- endif %}
+
 
 Response (should be a valid YAML, and nothing else):
 ```yaml

pr_agent/settings/pr_reviewer_prompts.toml

@@ -221,6 +221,59 @@ The PR code diff:
 ======
 
 
+{%- if duplicate_prompt_examples %}
+
+
+Example output:
+```yaml
+review:
+{%- if related_tickets %}
+  ticket_compliance_check:
+    - ticket_url: |
+        ...
+      ticket_requirements: |
+        ...
+      fully_compliant_requirements: |
+        ...
+      not_compliant_requirements: |
+        ...
+      overall_compliance_level: |
+        ...
+{%- endif %}
+{%- if require_estimate_effort_to_review %}
+  estimated_effort_to_review_[1-5]: |
+    3
+{%- endif %}
+{%- if require_score %}
+  score: 89
+{%- endif %}
+  relevant_tests: |
+    No
+  key_issues_to_review:
+    - relevant_file: |
+        ...
+      issue_header: |
+        ...
+      issue_content: |
+        ...
+      start_line: ...
+      end_line: ...
+    - ...
+  security_concerns: |
+    No
+{%- if require_can_be_split_review %}
+  can_be_split:
+  - relevant_files:
+    - ...
+    - ...
+    title: ...
+  - ...
+{%- endif %}
+```
+(replace '...' with the actual values)
+{%- endif %}
+
+
 Response (should be a valid YAML, and nothing else):
 ```yaml
 """

pr_agent/tools/pr_code_suggestions.py

@@ -81,6 +81,7 @@ def __init__(self, pr_url: str, cli_mode=False, args: list = None,
             "relevant_best_practices": "",
             "is_ai_metadata": get_settings().get("config.enable_ai_metadata", False),
             "focus_only_on_problems": get_settings().get("pr_code_suggestions.focus_only_on_problems", False),
+            'duplicate_prompt_examples': get_settings().config.get('duplicate_prompt_examples', False),
         }
         self.pr_code_suggestions_prompt_system = get_settings().pr_code_suggestions_prompt.system
 
@@ -830,7 +831,8 @@ async def self_reflect_on_suggestions(self,
                          "diff": patches_diff,
                          'num_code_suggestions': len(suggestion_list),
                          'prev_suggestions_str': prev_suggestions_str,
-                         "is_ai_metadata": get_settings().get("config.enable_ai_metadata", False)}
+                         "is_ai_metadata": get_settings().get("config.enable_ai_metadata", False),
+                         'duplicate_prompt_examples': get_settings().config.get('duplicate_prompt_examples', False)}
             environment = Environment(undefined=StrictUndefined)
 
             if dedicated_prompt:

pr_agent/tools/pr_description.py

@@ -71,7 +71,8 @@ def __init__(self, pr_url: str, args: list = None,
             "custom_labels_class": "",  # will be filled if necessary in 'set_custom_labels' function
             "enable_semantic_files_types": get_settings().pr_description.enable_semantic_files_types,
             "related_tickets": "",
-            "include_file_summary_changes": len(self.git_provider.get_diff_files()) <= self.COLLAPSIBLE_FILE_LIST_THRESHOLD
+            "include_file_summary_changes": len(self.git_provider.get_diff_files()) <= self.COLLAPSIBLE_FILE_LIST_THRESHOLD,
+            'duplicate_prompt_examples': get_settings().config.get('duplicate_prompt_examples', False),
         }
 
         self.user_description = self.git_provider.get_user_description()

pr_agent/tools/pr_reviewer.py

@@ -94,6 +94,7 @@ def __init__(self, pr_url: str, is_answer: bool = False, is_auto: bool = False,
             "enable_custom_labels": get_settings().config.enable_custom_labels,
             "is_ai_metadata":  get_settings().get("config.enable_ai_metadata", False),
             "related_tickets": get_settings().get('related_tickets', []),
+            'duplicate_prompt_examples': get_settings().config.get('duplicate_prompt_examples', False),
         }
 
         self.token_handler = TokenHandler(