Spaces:

MaduRox
/

Kalpana-API-GPU

Runtime error

App Files Files Community

Kalpana-API-GPU / app.py

Commit History

Aug 23

18:56
docs: remove studio link and add paradigm info to API tab

MaduRox

bdef46f
18:55
docs: add paradigm explanations under tabs

MaduRox

1307aaa
18:35
docs: add cURL equivalent and rename Swagger tab

MaduRox

25c3ffa
07:01
Fix Phase Attention Benchmark to simulate GQA 2 KV heads

MaduRox

81b6e37
06:56
Add torch.inference_mode to stop backward graph creeping

MaduRox

b48d71c
04:24
Fix Phase Attention Benchmark to simulate 24 layers

MaduRox

4c632bd

Aug 22

17:33
feat: Add True O(1) Phase Attention benchmark endpoint

MaduRox

67c5c69
16:05
feat: add real empirical benchmark harness with /run_benchmark endpoint

MaduRox

ae7c56f
09:24
docs: update Swagger tab with complete PowerShell, Python, JS, and cURL 2-step REST examples

MaduRox

095ce37
08:16
fix: configure clean chat template and hybrid RIF sliding window on T4 GPU

MaduRox

74aa433
06:56
fix: add system prompt, top_p, and repetition_penalty to prevent small model loop

MaduRox

ef5ecdb
06:26
fix: restore native Gradio 5 launcher for ZeroGPU prestart hooks

MaduRox

38dbeb1
06:19
feat: add FastAPI server with Live Swagger UI, OpenAI endpoints, and CORS

MaduRox

c8f0400
03:04
feat: upgrade holographic resolution to 2048 bands on NVIDIA A100

MaduRox

1e0f5c6

Aug 21

06:31
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

6018bf3
06:24
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

01c6ddb
06:07
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

c23e3c7
05:53
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

37d0686
05:20
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

b1909b4
05:18
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

efb9f26
05:16
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

1a668ac
05:13
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

ab09287
05:11
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

21e9521
05:02
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

c828501
05:00
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

2a33d72
04:58
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

7fc1739
04:55
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

2e4d3c6
04:37
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

c07d3e6

Aug 20

16:38
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

110959e
15:51
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

2f6a704
15:36
feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

MaduRox

b6b8b5d

Aug 12

05:50
fix: resolve AttributeError on llm.DEFAULT_MODELS in Gradio UI startup

Madusha Perera

b879636
05:48
fix: restore list_models optional request parameter

Madusha Perera

1469bd5
05:32
fix: update list_models request signature

Madusha Perera

13bbf0d
05:25
fix: define _RATE_WINDOWS dict to prevent server 500 error

Madusha Perera

607fdc9
05:21
fix: make _get_client_ip safe against None request

Madusha Perera

1b5b6b9
05:14
feat: GET /v1/models returns ONLY user-registered provider models

Madusha Perera

e197e6c

Aug 11

11:42
feat: enforce registered provider keys and remove HF serverless fallbacks

Madusha Perera

562cddf
10:59
fix: add 429 rate limit formatting and boilerplate chunk filtering for story retrieval

Madusha Perera

a0f0d11
10:54
fix: correct cost comparison formula based on full uncompressed document size vs compressed RIF tokens

Madusha Perera

6a02f22
10:43
feat: expand RIF retrieval to 1600 tokens and enable natural synthesis system prompt

Madusha Perera

1246a60
10:22
fix: auto-map gemini model to gemini-2.5-flash

Madusha Perera

3093778
10:08
fix: import time inside generate method

Madusha Perera

ae33d9b
10:07
fix: define t0 at top of generate method

Madusha Perera

3d0f535
10:05
fix: report registered provider key errors explicitly in chat response

Madusha Perera

66181bf
09:51
fix: route gemini and registered model IDs to custom provider API caller

Madusha Perera

aff6204
09:48
debug: expose last_err in RIF context fallback

Madusha Perera

9f7c7e5
09:46
fix: add browser User-Agent header to HF router requests for live LLM inference

Madusha Perera

048f6f9
09:35
fix: clear custom provider fallback on error and add /v1/providers/reset endpoint

Madusha Perera

a0b26e4
08:45
feat: add Google Gemini support and global latest_provider fallback

Madusha Perera

e311056

Commit History

Aug 23

docs: remove studio link and add paradigm info to API tab

docs: add paradigm explanations under tabs

docs: add cURL equivalent and rename Swagger tab

Fix Phase Attention Benchmark to simulate GQA 2 KV heads

Add torch.inference_mode to stop backward graph creeping

Fix Phase Attention Benchmark to simulate 24 layers

Aug 22

feat: Add True O(1) Phase Attention benchmark endpoint

feat: add real empirical benchmark harness with /run_benchmark endpoint

docs: update Swagger tab with complete PowerShell, Python, JS, and cURL 2-step REST examples

fix: configure clean chat template and hybrid RIF sliding window on T4 GPU

fix: add system prompt, top_p, and repetition_penalty to prevent small model loop

fix: restore native Gradio 5 launcher for ZeroGPU prestart hooks

feat: add FastAPI server with Live Swagger UI, OpenAI endpoints, and CORS

feat: upgrade holographic resolution to 2048 bands on NVIDIA A100

Aug 21

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

Aug 20

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

feat: integrate KalpanaDynamicCache into all 24 layers of Qwen2.5 on ZeroGPU

Aug 12

fix: resolve AttributeError on llm.DEFAULT_MODELS in Gradio UI startup

fix: restore list_models optional request parameter

fix: update list_models request signature

fix: define _RATE_WINDOWS dict to prevent server 500 error

fix: make _get_client_ip safe against None request

feat: GET /v1/models returns ONLY user-registered provider models

Aug 11

feat: enforce registered provider keys and remove HF serverless fallbacks

fix: add 429 rate limit formatting and boilerplate chunk filtering for story retrieval

fix: correct cost comparison formula based on full uncompressed document size vs compressed RIF tokens

feat: expand RIF retrieval to 1600 tokens and enable natural synthesis system prompt

fix: auto-map gemini model to gemini-2.5-flash

fix: import time inside generate method

fix: define t0 at top of generate method

fix: report registered provider key errors explicitly in chat response

fix: route gemini and registered model IDs to custom provider API caller

debug: expose last_err in RIF context fallback

fix: add browser User-Agent header to HF router requests for live LLM inference

fix: clear custom provider fallback on error and add /v1/providers/reset endpoint

feat: add Google Gemini support and global latest_provider fallback