Skip to content

Commit 8c1370a

Browse files
lchoquelclaude
andcommitted
Update LLM deck defaults, fix object list e2e test, add bedrock token auth changelog
Bump best-claude to opus-4.7, swap small-vision/creative to gemini-3.0-flash-preview, add cheap presets for writing/retrieval/engineering, fix bedrock_access_variant default back to aws_access, and reduce moodboard prompt test from 5 to 3 items to match the updated .mthds fixture. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
1 parent faa518a commit 8c1370a

5 files changed

Lines changed: 18 additions & 16 deletions

File tree

CHANGELOG.md

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -9,6 +9,7 @@
99
- **`TEMPERATURE_UNSUPPORTED` constraint**: New listed constraint for models that reject sampling parameters entirely, checked in both Anthropic and OpenAI completions workers
1010
- **claude-4.6-sonnet model**: Registered on anthropic, bedrock, and gateway backends
1111
- **LLM deck cheap presets**: Added cheap variants for writing-factual, retrieval, and engineering-code presets, with retrieval tiers from `gemini-2.5-flash-lite` to `claude-4.7-opus`
12+
- **Bedrock bearer token authentication**: New `bedrock_access_variant` config option supports `"bedrock_token"` auth using `AWS_BEARER_TOKEN_BEDROCK` env var, alongside the existing `"aws_access"` method (default)
1213

1314
### Changed
1415

pipelex/kit/configs/inference/deck/1_llm_deck.toml

Lines changed: 12 additions & 10 deletions
Original file line numberDiff line numberDiff line change
@@ -28,7 +28,7 @@ for_object = "@default-general"
2828

2929
[llm.aliases]
3030
best-gpt = "gpt-5.2"
31-
best-claude = "claude-4.6-opus"
31+
best-claude = "claude-4.7-opus"
3232
best-gemini = "gemini-3.0-pro"
3333
best-mistral = "mistral-large"
3434

@@ -41,8 +41,8 @@ default-large-context-code = "gemini-3.0-pro"
4141
default-large-context-text = "gemini-2.5-flash"
4242
default-small = "gpt-4o-mini"
4343
default-small-structured = "gpt-4o-mini"
44-
default-small-vision = "gemini-2.5-flash-lite"
45-
default-small-creative = "gemini-2.5-flash-lite"
44+
default-small-vision = "gemini-3.0-flash-preview"
45+
default-small-creative = "gemini-3.0-flash-preview"
4646

4747
####################################################################################################
4848
# LLM Presets
@@ -52,34 +52,36 @@ default-small-creative = "gemini-2.5-flash-lite"
5252

5353
# Writing
5454
writing-factual = { model = "@default-premium", temperature = 0.1, description = "Factual writing with high accuracy" }
55+
writing-factual-cheap = { model = "@default-small", temperature = 0.1, description = "Cheap factual writing with high accuracy" }
5556
writing-creative = { model = "@default-premium", temperature = 0.9, description = "Creative writing with high variability" }
57+
writing-creative-cheap = { model = "@default-small-creative", temperature = 0.9, description = "Cheap creative writing with high variability" }
5658

5759
# Retrieval
5860
retrieval = { model = "@default-large-context-text", temperature = 0.1, description = "Data retrieval from large text corpora" }
61+
retrieval-cheap = { model = "gemini-2.5-flash-lite", temperature = 0.1, description = "Cheap data retrieval from large text corpora" }
62+
retrieval-premium = { model = "claude-4.7-opus", temperature = 0.1, description = "Premium data retrieval with highest accuracy" }
5963

6064
# Engineering
6165
engineering-structured = { model = "@default-premium-structured", temperature = 0.2, description = "Structured engineering output (JSON, schemas)" }
6266
engineering-code = { model = "@default-premium", temperature = 0.1, description = "Code generation and analysis" }
67+
engineering-code-cheap = { model = "claude-4.6-sonnet", temperature = 0.1, description = "Cheap code generation and analysis" }
68+
engineering-code-cheaper = { model = "claude-4.5-haiku", temperature = 0.1, description = "Cheapest code generation and analysis" }
6369
engineering-codebase-analysis = { model = "@best-gemini", temperature = 0.1, description = "Large codebase analysis" }
6470

6571
# Vision
66-
vision = { model = "@default-premium-vision", temperature = 0.5, description = "Vision language model for understanding images" }
67-
vision-cheap = { model = "@default-small-vision", temperature = 0.5, description = "Budget vision model for simple image tasks" }
72+
vision = { model = "@default-premium-vision", temperature = 0.5, description = "Understanding and interpreting images" }
73+
vision-cheap = { model = "@default-small-vision", temperature = 0.5, description = "Cheap image understanding for simple tasks" }
6874
vision-diagram = { model = "@default-premium-vision", temperature = 0.3, description = "Diagram and chart interpretation" }
6975
vision-table = { model = "@default-premium-vision", temperature = 0.3, description = "Table extraction from images" }
7076

7177
# Image generation prompting
7278
img-gen-prompting = { model = "@default-premium", temperature = 0.5, description = "Crafting image generation prompts" }
73-
img-gen-prompting-cheap = { model = "@default-small", temperature = 0.5, description = "Budget image prompt generation" }
79+
img-gen-prompting-cheap = { model = "@default-small-creative", temperature = 0.5, description = "Cheap image prompt generation" }
7480

7581
# Reasoning
7682
deep-analysis = { model = "@default-premium", temperature = 0.1, reasoning_effort = "high", description = "Deep reasoning and analysis" }
7783
quick-reasoning = { model = "@default-premium", temperature = 0.3, reasoning_effort = "low", description = "Quick reasoning for simple tasks" }
7884

79-
# Builder (isolated presets for the pipeline builder)
80-
pipe-builder-engineering = { model = "claude-4.6-opus", temperature = 0.2, description = "Builder: structured engineering output" }
81-
pipe-builder-img-gen-prompting = { model = "claude-4.5-sonnet", temperature = 0.7, description = "Builder: crafting image generation prompts" }
82-
8385
# Testing
8486
testing-text = { model = "@default-small", temperature = 0.5, description = "Testing preset for text generation" }
8587
testing-structured = { model = "@default-small-structured", temperature = 0.1, description = "Testing preset for structured output" }

pipelex/pipelex.toml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -117,7 +117,7 @@ keywords_to_hilight = []
117117

118118
[pipelex.aws_config]
119119
api_key_method = "env"
120-
bedrock_access_variant = "bedrock_token"
120+
bedrock_access_variant = "aws_access"
121121

122122
####################################################################################################
123123
# Cogt inference config

tests/e2e/pipelex/pipes/pipe_operators/pipe_llm/pipe_llm_object_list.mthds

Lines changed: 2 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -14,13 +14,12 @@ inputs = { inspiration = "Text" }
1414
output = "MoodboardPrompt[]"
1515
model = "@default-small-creative"
1616
prompt = """
17-
You are a fashion art director creating a moodboard. Given the following fashion inspiration, generate exactly 5 distinct image generation prompts. Each prompt should capture a different visual facet of the aesthetic:
17+
You are a fashion art director creating a moodboard. Given the following fashion inspiration, generate exactly 3 distinct image generation prompts.
18+
Each prompt should capture a different visual facet of the aesthetic:
1819

1920
1. A hero outfit or look
2021
2. A texture or fabric close-up
2122
3. A color palette scene
22-
4. A setting or environment that evokes the mood
23-
5. An accessory or detail shot
2423

2524
Fashion inspiration:
2625
@inspiration

tests/e2e/pipelex/pipes/pipe_operators/pipe_llm/test_pipe_llm_object_list.py

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -11,7 +11,7 @@
1111
@pytest.mark.llm
1212
@pytest.mark.inference
1313
@pytest.mark.dry_runnable
14-
@pytest.mark.asyncio(loop_scope="class")
14+
@pytest.mark.asyncio
1515
class TestPipeLLMObjectList:
1616
async def test_craft_prompts(self, pipe_run_mode: PipeRunMode) -> None:
1717
"""Test a PipeLLM pipe that generates a list of structured MoodboardPrompt objects."""
@@ -29,6 +29,6 @@ async def test_craft_prompts(self, pipe_run_mode: PipeRunMode) -> None:
2929
assert pipeline_response.pipe_output.main_stuff is not None
3030

3131
items = pipeline_response.pipe_output.main_stuff_as_list(item_type=TextContent)
32-
assert len(items) == 5
32+
assert len(items) == 3
3333

3434
pretty_print(items, title="Moodboard prompts")

0 commit comments

Comments
 (0)