python-fastllm 0.0.55__py3-none-any.whl → 0.0.57__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- fastllm/__init__.py +1 -1
- fastllm/anthropic.py +3 -2
- fastllm/gemini.py +1 -1
- fastllm/model_prices.json +58559 -0
- fastllm/openai_responses.py +1 -2
- fastllm/types.py +3 -3
- {python_fastllm-0.0.55.dist-info → python_fastllm-0.0.57.dist-info}/METADATA +1 -1
- python_fastllm-0.0.57.dist-info/RECORD +18 -0
- python_fastllm-0.0.55.dist-info/RECORD +0 -17
- {python_fastllm-0.0.55.dist-info → python_fastllm-0.0.57.dist-info}/WHEEL +0 -0
- {python_fastllm-0.0.55.dist-info → python_fastllm-0.0.57.dist-info}/entry_points.txt +0 -0
- {python_fastllm-0.0.55.dist-info → python_fastllm-0.0.57.dist-info}/top_level.txt +0 -0
fastllm/__init__.py
CHANGED
fastllm/anthropic.py
CHANGED
|
@@ -44,7 +44,8 @@ def norm_usage(resp):
|
|
|
44
44
|
pt = int(usg.get("input_tokens", 0) or 0) + cached + cache_creation
|
|
45
45
|
ct = int(usg.get("output_tokens", 0) or 0)
|
|
46
46
|
return Usage(prompt_tokens=pt, completion_tokens=ct, total_tokens=pt + ct,
|
|
47
|
-
cached_tokens=cached, cache_creation_tokens=cache_creation, reasoning_tokens=0, raw=usg
|
|
47
|
+
cached_tokens=cached, cache_creation_tokens=cache_creation, reasoning_tokens=0, raw=usg,
|
|
48
|
+
search_queries=int(nested_idx(usg, 'server_tool_use', 'web_search_requests') or 0))
|
|
48
49
|
|
|
49
50
|
def finalize_usage(usg, parts):
|
|
50
51
|
"Split reasoning tokens out of `output_tokens`: the reported `thinking_tokens`, else an estimate from the thinking text."
|
|
@@ -53,7 +54,7 @@ def finalize_usage(usg, parts):
|
|
|
53
54
|
ct = int(usg.raw.get('output_tokens', usg.completion_tokens) or 0)
|
|
54
55
|
rt = int(nested_idx(usg.raw, 'output_tokens_details', 'thinking_tokens') or 0) or (min(int(len(rc.split())*1.5), ct) if rc else 0)
|
|
55
56
|
return Usage(prompt_tokens=usg.prompt_tokens, completion_tokens=ct-rt, total_tokens=usg.prompt_tokens+ct,
|
|
56
|
-
cached_tokens=usg.cached_tokens, cache_creation_tokens=usg.cache_creation_tokens, reasoning_tokens=rt, raw=usg.raw)
|
|
57
|
+
cached_tokens=usg.cached_tokens, cache_creation_tokens=usg.cache_creation_tokens, reasoning_tokens=rt, search_queries=usg.search_queries, raw=usg.raw)
|
|
57
58
|
|
|
58
59
|
# %% ../nbs/04_anthropic.ipynb #7a8b1f8f
|
|
59
60
|
def norm_finish(resp, tcs=None):
|
fastllm/gemini.py
CHANGED
|
@@ -67,7 +67,7 @@ def norm_usage(resp):
|
|
|
67
67
|
if call := grounding(cand)[1]: stu['google_search'] = len(call.arguments['queries'] or [1])
|
|
68
68
|
if stu: usg['server_tool_use'] = stu
|
|
69
69
|
return Usage(prompt_tokens=pt, completion_tokens=ct, total_tokens=tt,
|
|
70
|
-
cached_tokens=cached, reasoning_tokens=reasoning, raw=usg)
|
|
70
|
+
cached_tokens=cached, reasoning_tokens=reasoning, search_queries=stu.get('google_search', 0), raw=usg)
|
|
71
71
|
|
|
72
72
|
# %% ../nbs/05_gemini.ipynb #7a8b1f8f
|
|
73
73
|
def norm_finish(resp, tcs=None):
|