Merge branch 'main' into feature/add-MyScale-to-retrieve

Author: arnavsinghvi11

README.md

@@ -72,12 +72,11 @@ Or open our intro notebook in Google Colab: [<img align="center" src="https://co
 
 By default, DSPy installs the latest `openai` from pip. However, if you install old version before OpenAI changed their API `openai~=0.28.1`, the library will use that just fine. Both are supported.
 
+For the optional (alphabetically sorted) [Chromadb](https://github.com/chroma-core/chroma), [Groq](https://github.com/groq/groq-python), [Marqo](https://github.com/marqo-ai/marqo), [Milvus](https://github.com/milvus-io/milvus), [MongoDB](https://www.mongodb.com), [MyScaleDB](https://github.com/myscale/myscaledb), Pinecone, [Qdrant](https://github.com/qdrant/qdrant), [Snowflake](https://github.com/snowflakedb/snowpark-python), or [Weaviate](https://github.com/weaviate/weaviate) retrieval integration(s), include the extra(s) below:
 
-For the optional (alphabetically sorted) [Chromadb](https://github.com/chroma-core/chroma), [Marqo](https://github.com/marqo-ai/marqo), [Milvus](https://github.com/milvus-io/milvus), [MyScaleDB](https://github.com/myscale/myscaledb), MongoDB, [MyScaleDB](https://github.com/myscale/myscaledb), Pinecone, [Qdrant](https://github.com/qdrant/qdrant), or [Weaviate](https://github.com/weaviate/weaviate) retrieval integration(s), include the extra(s) below:
-
 ```
-pip install dspy-ai[chromadb]  # or [marqo] or [milvus] or [myscale] or [mongodb] or [myscale] or [pinecone] or [qdrant] or [weaviate] 
-``````
+pip install dspy-ai[chromadb] # or [groq] or [marqo] or [milvus] or [mongodb] or [myscale] or [pinecone] or [qdrant] or [snowflake] or [weaviate]
+```
 
 ## 2) Documentation
 

docs/api/functional/dspy_cot.md

@@ -8,7 +8,7 @@ sidebar_position: 4
 
 #### `def cot(func) -> dspy.Module`
 
-The `@cot` decorator is used to create a Chain of Thoughts module based on the provided function. It automatically generates a `dspy.TypedPredictor` and from the function's type annotations and docstring. Similar to predictor, but adds a "Reasoning" output field to capture the model's step-by-step thinking.
+The `@cot` decorator is used to create a Chain of Thoughts module based on the provided function. It automatically generates a `dspy.TypedPredictor` from the function's type annotations and docstring. Similar to predictor, but adds a "Reasoning" output field to capture the model's step-by-step thinking.
 
 * **Input**: Function with input parameters and return type annotation.
 * **Output**: A dspy.Module instance capable of making predictions.
@@ -27,4 +27,4 @@ def generate_answer(self, context: list[str], question) -> str:
     pass
 
 generate_answer(context=context, question=question)
-```
+```

docs/api/local_language_model_clients/Ollama.md

@@ -29,7 +29,7 @@ Here is the list of other models you can download:
 
 Run model: `ollama run`
 
-You can test a model by running the model with the `ollama run` command.
+You need to start the model server with the `ollama run` command.
 
 ```bash
 # run mistral

docs/docs/cheatsheet.md

@@ -280,7 +280,7 @@ fewshot_optimizer = BootstrapFewShot(metric=your_defined_metric, max_bootstrappe
 your_dspy_program_compiled = fewshot_optimizer.compile(student = your_dspy_program, trainset=trainset)
 ```
 
-#### Compiling a compiled program - bootstrapping a bootstraped program
+#### Compiling a compiled program - bootstrapping a bootstrapped program
 
 ```python
 your_dspy_program_compiledx2 = teleprompter.compile(
@@ -363,7 +363,7 @@ from dspy.teleprompt import COPRO
 
 eval_kwargs = dict(num_threads=16, display_progress=True, display_table=0)
 
-copro_teleprompter = COPRO(prompt_model=model_to_generate_prompts, task_model=model_that_solves_task, metric=your_defined_metric, breadth=num_new_prompts_generated, depth=times_to_generate_prompts, init_temperature=prompt_generation_temperature, verbose=False, log_dir=logging_directory)
+copro_teleprompter = COPRO(prompt_model=model_to_generate_prompts, metric=your_defined_metric, breadth=num_new_prompts_generated, depth=times_to_generate_prompts, init_temperature=prompt_generation_temperature, verbose=False)
 
 compiled_program_optimized_signature = copro_teleprompter.compile(your_dspy_program, trainset=trainset, eval_kwargs=eval_kwargs)
 ```

docs/docs/faqs.md

@@ -81,7 +81,7 @@ Exporting DSPy programs is simply saving them as highlighted above!
 
 - **How do I search my own data?**
 
-Open source libraries such as [RAGautouille](https://github.com/bclavie/ragatouille) enable you to search for your own data through advanced retrieval models like ColBERT with tools to embdeed and index documents. Feel free to integrate such libraries to create searchable datasets while developing your DSPy programs!
+Open source libraries such as [RAGautouille](https://github.com/bclavie/ragatouille) enable you to search for your own data through advanced retrieval models like ColBERT with tools to embed and index documents. Feel free to integrate such libraries to create searchable datasets while developing your DSPy programs!
 
 - **How do I turn off the cache? How do I export the cache?**
 

dsp/modules/cohere.py

@@ -63,7 +63,7 @@ def __init__(
         self.kwargs = {
             "model": model,
             "temperature": 0.0,
-            "max_tokens": 150,
+            "max_tokens": 2000,
             "p": 1,
             "num_generations": 1,
             **kwargs,

dsp/modules/lm.py

@@ -46,13 +46,14 @@ def inspect_history(self, n: int = 1, skip: int = 0):
             prompt = x["prompt"]
 
             if prompt != last_prompt:
-                if (
-                    provider == "clarifai"
-                    or provider == "google"
-                    or provider == "groq"
-                    or provider == "Bedrock"
-                    or provider == "Sagemaker"
-                    or provider == "premai"
+                if provider in (
+                    "clarifai",
+                    "cloudflare"
+                    "google",
+                    "groq",
+                    "Bedrock",
+                    "Sagemaker",
+                    "premai",
                 ):
                     printed.append((prompt, x["response"]))
                 elif provider == "anthropic":
@@ -66,8 +67,6 @@ def inspect_history(self, n: int = 1, skip: int = 0):
                     printed.append((prompt, x["response"].text))
                 elif provider == "mistral":
                     printed.append((prompt, x["response"].choices))
-                elif provider == "cloudflare":
-                    printed.append((prompt, [x["response"]]))
                 elif provider == "ibm":
                     printed.append((prompt, x))
                 else:
@@ -87,12 +86,10 @@ def inspect_history(self, n: int = 1, skip: int = 0):
             printing_value += prompt
 
             text = ""
-            if provider == "cohere" or provider == "Bedrock" or provider == "Sagemaker":
+            if provider in ("cohere", "Bedrock", "Sagemaker", "clarifai", "claude", "ibm", "premai"):
                 text = choices
             elif provider == "openai" or provider == "ollama":
                 text = " " + self._get_choice_text(choices[0]).strip()
-            elif provider == "clarifai" or provider == "claude":
-                text = choices
             elif provider == "groq":
                 text = " " + choices
             elif provider == "google":
@@ -101,8 +98,6 @@ def inspect_history(self, n: int = 1, skip: int = 0):
                 text = choices[0].message.content
             elif provider == "cloudflare":
                 text = choices[0]
-            elif provider == "ibm" or provider == "premai":
-                text = choices
             else:
                 text = choices[0]["text"]
             printing_value += self.print_green(text, end="")

dspy/predict/multi_chain_comparison.py

@@ -42,7 +42,7 @@ def forward(self, completions, **kwargs):
                 f"«I'm trying to {rationale} I'm not sure but my prediction is {answer}»",
             )
 
-        assert len(attempts) == self.M, len(attempts)
+        assert len(attempts) == self.M, f"The number of attempts ({len(attempts)}) doesn't match the expected number M ({self.M}). Please set the correct value for M when initializing MultiChainComparison."
 
         kwargs = {
             **{

dspy/teleprompt/bootstrap.py

@@ -112,10 +112,17 @@ def _prepare_predictor_mappings(self):
 
         for (name1, predictor1), (name2, predictor2) in zip(student.named_predictors(), teacher.named_predictors()):
             assert name1 == name2, "Student and teacher must have the same program structure."
-            assert predictor1.signature.equals(
-                predictor2.signature,
-            ), (f"Student and teacher must have the same signatures. "
-                f"{type(predictor1.signature)} != {type(predictor2.signature)}"
+            if hasattr(predictor1.signature, "equals"):
+                assert predictor1.signature.equals(
+                    predictor2.signature,
+                ), (f"Student and teacher must have the same signatures. "
+                    f"{type(predictor1.signature)} != {type(predictor2.signature)}"
+                    )
+            else:
+                # fallback in case if .equals is not implemented (e.g. dsp.Prompt)
+                assert predictor1.signature == predictor2.signature, (
+                    f"Student and teacher must have the same signatures. "
+                    f"{type(predictor1.signature)} != {type(predictor2.signature)}"
                 )
             assert id(predictor1) != id(predictor2), "Student and teacher must be different objects."
 

setup.py

@@ -37,6 +37,7 @@
         "fastembed": ["fastembed"],
         "google-vertex-ai": ["google-cloud-aiplatform==1.43.0"],
         "myscale":["clickhouse-connect"],
+        "groq": ["groq~=0.8.0"],
     },	
     classifiers=[	
         "Development Status :: 3 - Alpha",