mirror of
https://github.com/BerriAI/litellm.git
synced 2026-09-20 00:11:50 +00:00
test: remove test files left empty by the cleanup
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
This commit is contained in:
kerry
parent
c94a4f8696
commit
e64371062c
5 changed files with 0 additions and 148 deletions
|
|
@ -1,54 +0,0 @@
|
|||
"""
|
||||
Regression test for Fireworks Kimi K2.5 / K2.6 / K2.7 context and output limits.
|
||||
|
||||
Fireworks publishes a 262144-token context window for every Kimi K2.5, K2.6 and
|
||||
K2.7 model, but caps generation well below that. A previous bulk edit had flattened
|
||||
max_output_tokens/max_tokens to 262144 (equal to the context window), which let the
|
||||
pre-call context-window check admit requests asking for a full 262144-token
|
||||
completion that Fireworks then rejects. These assertions pin the corrected per-alias
|
||||
limits so a future bulk edit can't silently flatten them again.
|
||||
"""
|
||||
|
||||
import json
|
||||
from importlib.resources import files
|
||||
|
||||
import pytest
|
||||
|
||||
CONTEXT_WINDOW = 262144
|
||||
OUTPUT_LIMIT = 32768
|
||||
|
||||
KIMI_ALIASES = (
|
||||
"fireworks_ai/kimi-k2p5",
|
||||
"fireworks_ai/kimi-k2p6",
|
||||
"fireworks_ai/kimi-k2p6-fast",
|
||||
"fireworks_ai/kimi-k2p7-code",
|
||||
"fireworks_ai/kimi-k2p7-code-fast",
|
||||
"fireworks_ai/accounts/fireworks/models/kimi-k2p5",
|
||||
"fireworks_ai/accounts/fireworks/models/kimi-k2p6",
|
||||
"fireworks_ai/accounts/fireworks/models/kimi-k2p7-code",
|
||||
"fireworks_ai/accounts/fireworks/routers/kimi-k2p6-fast",
|
||||
"fireworks_ai/accounts/fireworks/routers/kimi-k2p7-code-fast",
|
||||
)
|
||||
|
||||
|
||||
@pytest.fixture(scope="module")
|
||||
def use_local_model_cost_map():
|
||||
monkeypatch = pytest.MonkeyPatch()
|
||||
monkeypatch.setenv("LITELLM_LOCAL_MODEL_COST_MAP", "True")
|
||||
|
||||
import litellm
|
||||
from litellm.utils import _invalidate_model_cost_lowercase_map
|
||||
|
||||
original_model_cost = litellm.model_cost
|
||||
litellm.model_cost = json.loads(
|
||||
files("litellm").joinpath("model_prices_and_context_window_backup.json").read_text(encoding="utf-8")
|
||||
)
|
||||
litellm.get_model_info.cache_clear()
|
||||
_invalidate_model_cost_lowercase_map()
|
||||
try:
|
||||
yield litellm
|
||||
finally:
|
||||
litellm.model_cost = original_model_cost
|
||||
litellm.get_model_info.cache_clear()
|
||||
_invalidate_model_cost_lowercase_map()
|
||||
monkeypatch.undo()
|
||||
|
|
@ -1,40 +0,0 @@
|
|||
"""
|
||||
Cost tests for Mistral OCR models against the real litellm cost map
|
||||
(no monkeypatching of get_model_info). These regress the pricing entries
|
||||
for mistral-ocr-4-0 and mistral-ocr-latest, which now both resolve to
|
||||
OCR 4 at $4 / 1000 pages.
|
||||
"""
|
||||
|
||||
from pathlib import Path
|
||||
|
||||
from litellm.llms.base_llm.ocr.transformation import OCRPage, OCRResponse, OCRUsageInfo
|
||||
|
||||
OCR4_COST_PER_PAGE = 0.004
|
||||
OCR4_ANNOTATION_COST_PER_PAGE = 0.005
|
||||
|
||||
REPO_ROOT = Path(__file__).parents[5]
|
||||
MAIN_COST_MAP = REPO_ROOT / "model_prices_and_context_window.json"
|
||||
BACKUP_COST_MAP = REPO_ROOT / "litellm" / "model_prices_and_context_window_backup.json"
|
||||
|
||||
OCR3_MODEL = "mistral/mistral-ocr-2512"
|
||||
OCR3_COST_PER_PAGE = 0.002
|
||||
OCR3_ANNOTATION_COST_PER_PAGE = 0.003
|
||||
|
||||
AZURE_DOC_AI_MODEL = "azure_ai/mistral-document-ai-2512"
|
||||
AZURE_DOC_AI_COST_PER_PAGE = 0.003
|
||||
|
||||
|
||||
def _ocr_response(model: str, pages_processed: int) -> OCRResponse:
|
||||
return OCRResponse(
|
||||
pages=[OCRPage(index=i, markdown=f"page {i}") for i in range(pages_processed)],
|
||||
model=model,
|
||||
usage_info=OCRUsageInfo(pages_processed=pages_processed),
|
||||
)
|
||||
|
||||
|
||||
def _annotated_ocr_response(model: str, pages_processed: int | None, annotation_pages: int) -> OCRResponse:
|
||||
return OCRResponse(
|
||||
pages=[],
|
||||
model=model,
|
||||
usage_info=OCRUsageInfo(pages_processed=pages_processed, pages_processed_annotation=annotation_pages),
|
||||
)
|
||||
|
|
@ -1 +0,0 @@
|
|||
|
||||
|
|
@ -1,53 +0,0 @@
|
|||
"""
|
||||
Regression test: ``command-r7b-12-2024`` had its input/output per-token
|
||||
costs transposed in the model-cost maps (input=1.5e-07 / output=3.75e-08),
|
||||
even though Cohere publishes $0.0375/1M input and $0.15/1M output, i.e.
|
||||
output is ~4x input like every other ``command-r`` entry.
|
||||
|
||||
These tests pin the corrected values in both the primary price map and the
|
||||
``litellm/`` backup, and verify ``get_model_info`` surfaces them, so the
|
||||
swap cannot silently regress.
|
||||
"""
|
||||
|
||||
import json
|
||||
import os
|
||||
|
||||
import litellm
|
||||
|
||||
MODEL = "command-r7b-12-2024"
|
||||
EXPECTED_INPUT_COST = 3.75e-08
|
||||
EXPECTED_OUTPUT_COST = 1.5e-07
|
||||
|
||||
|
||||
def _load_json(path: str) -> dict:
|
||||
with open(path, encoding="utf-8") as f:
|
||||
return json.load(f)
|
||||
|
||||
|
||||
def _backup_path() -> str:
|
||||
return os.path.join(
|
||||
os.path.dirname(litellm.__file__),
|
||||
"model_prices_and_context_window_backup.json",
|
||||
)
|
||||
|
||||
|
||||
def _main_path() -> str:
|
||||
# This test lives at ``tests/test_litellm/``; the primary price map sits at
|
||||
# the repo root, two directories up. Resolve it relative to this file so the
|
||||
# test works regardless of where ``litellm`` itself is installed (e.g. a pip
|
||||
# install into site-packages).
|
||||
return os.path.join(
|
||||
os.path.dirname(__file__),
|
||||
"..",
|
||||
"..",
|
||||
"model_prices_and_context_window.json",
|
||||
)
|
||||
|
||||
|
||||
class TestCommandR7bPricingData:
|
||||
"""The JSON price maps must carry Cohere's published costs, with output
|
||||
more expensive than input."""
|
||||
|
||||
|
||||
class TestCommandR7bPricingModelInfo:
|
||||
"""``get_model_info`` must report the corrected, un-swapped costs."""
|
||||
Loading…
Add table
Reference in a new issue