{ "evidenceVersion": "local-ollama-tps-sweep-v1", "capturedAt": "2026-05-28T05:13:10.545Z", "evidenceUrl": "https://www.runlocalai.co/benchmarks/evidence/local-ollama-rtx-5080-2026-05-28-turkish-sweep/turkcell-llm-7b/evidence.json", "hardwareSlug": "rtx-5080", "hardwareLabel": "rtx-5080-16gb", "gpu": "NVIDIA GeForce RTX 5080 16GB", "runtime": "ollama-local-api", "scenario": "single-stream", "command": "POST http://localhost:11434/api/generate model=RefinedNeuro/Turkcell-LLM-7b-v1:latest temperature=0 seed=42 num_ctx=2048 num_predict=512", "contextSize": 2048, "numPredict": 512, "seed": 42, "temperature": 0, "promptSha256": "843f1934769f40e6972f520eafd20f336ef4fa86d7f50530aa03477d411084fa", "protocolNote": "Batch=1, 2048 context, 512 generated tokens, consecutive warm-model runs. Generation TPS uses Ollama eval_count/eval_duration.", "outputHash": "12c8d822940557b9a8381fcd5594d09e62b5ca1bcc036192447ba812eb8d688d", "sourceFileSha256": "897d20358d0a41bb640d93209207220980320a91618c1f3a18a382dbbe22e414", "sourceResultSha256": "6ffc8a6782bb9e94aa30524c89da9867d4caa5829a043f932e661c2d3b91d037", "artifactHashSha256": "1025398db148efe5695f432fbd7bed17af4d13623a5bcf73a929c39c101d8ab0", "logHashSha256": "1025398db148efe5695f432fbd7bed17af4d13623a5bcf73a929c39c101d8ab0", "logHashNote": "Structured Ollama API sweep hash. No separate verbose text logs were captured for this sweep.", "modelTag": "RefinedNeuro/Turkcell-LLM-7b-v1:latest", "catalogSlug": "turkcell-llm-7b", "family": "llama", "parameterSize": "7.4B", "quantizationLevel": "Q4_K_M", "modelSizeBytes": 4461475966, "digest": "50f61cb255fad048b2e56684390e8397b69341f90644289d6182671ab197167d", "isTurkishModel": true, "scoresGenerationTps": [ 146.35, 145.77, 144.04, 145.08, 144.99 ], "statsGenerationTps": { "min": 144.04, "max": 146.35, "median": 145.08, "p5": 144.23, "p95": 146.234, "mean": 145.246, "stdev": 0.7798615261698718, "cvPct": 0.5369246149084118, "n": 5 }, "runsCaptured": 5, "runStatsJson": { "runs": [ { "runIndex": 1, "tokPerSec": 146.35, "totalTokPerSec": 136.48, "ttftMs": null, "prefillTokPerSec": 3231.2586995426527, "promptTokens": 65, "responseTokens": 494, "promptEvalMs": 20.116, "evalMs": 3375.5207, "totalMs": 3619.6559, "wallMs": 3623, "doneReason": "stop", "responseHash": "ac73f2fbd8e24088a967d02ab3534ba60916e20872c2be5a1e0fe9358d7854f8" }, { "runIndex": 2, "tokPerSec": 145.77, "totalTokPerSec": 135.46, "ttftMs": null, "prefillTokPerSec": 7452.419169915157, "promptTokens": 65, "responseTokens": 500, "promptEvalMs": 8.722, "evalMs": 3429.9851, "totalMs": 3691.0972, "wallMs": 3694, "doneReason": "stop", "responseHash": "f8ada0608c9ac3ac7b205e8300831ddf686a841afa34c3dc216085fb8d09035e" }, { "runIndex": 3, "tokPerSec": 144.04, "totalTokPerSec": 135.1, "ttftMs": null, "prefillTokPerSec": 7062.146892655367, "promptTokens": 65, "responseTokens": 500, "promptEvalMs": 9.204, "evalMs": 3471.1786, "totalMs": 3700.9201, "wallMs": 3704, "doneReason": "stop", "responseHash": "f8ada0608c9ac3ac7b205e8300831ddf686a841afa34c3dc216085fb8d09035e" }, { "runIndex": 4, "tokPerSec": 145.08, "totalTokPerSec": 135.9, "ttftMs": null, "prefillTokPerSec": 7548.922826781255, "promptTokens": 65, "responseTokens": 500, "promptEvalMs": 8.6105, "evalMs": 3446.3658, "totalMs": 3679.2439, "wallMs": 3682, "doneReason": "stop", "responseHash": "f8ada0608c9ac3ac7b205e8300831ddf686a841afa34c3dc216085fb8d09035e" }, { "runIndex": 5, "tokPerSec": 144.99, "totalTokPerSec": 135.65, "ttftMs": null, "prefillTokPerSec": 7541.828139140927, "promptTokens": 65, "responseTokens": 500, "promptEvalMs": 8.6186, "evalMs": 3448.5578, "totalMs": 3686.0435, "wallMs": 3690, "doneReason": "stop", "responseHash": "f8ada0608c9ac3ac7b205e8300831ddf686a841afa34c3dc216085fb8d09035e" } ], "coldStart": null, "steady": { "min": 144.04, "max": 146.35, "median": 145.08, "p5": 144.23, "p95": 146.234, "mean": 145.246, "stdev": 0.7798615261698718, "cvPct": 0.5369246149084118, "n": 5 }, "warm": { "min": 144.04, "max": 146.35, "median": 145.08, "p5": 144.23, "p95": 146.234, "mean": 145.246, "stdev": 0.7798615261698718, "cvPct": 0.5369246149084118, "n": 5 }, "power": null, "accuracy": null, "env": { "hardware": "rtx-5080-16gb", "hardwareSlug": "rtx-5080", "gpu": "NVIDIA GeForce RTX 5080 16GB", "runtime": "ollama-local-api" }, "protocol": { "benchmark": "yerelai-local-ollama-tps-sweep", "dimension": "tps", "sampleCount": 5, "contextLength": 2048, "numPredict": 512, "seed": 42, "temperature": 0, "promptSha256": "843f1934769f40e6972f520eafd20f336ef4fa86d7f50530aa03477d411084fa", "protocolNote": "Batch=1, 2048 context, 512 generated tokens, consecutive warm-model runs. Generation TPS uses Ollama eval_count/eval_duration.", "warmModelRuns": true, "batchSize": 1, "concurrency": 1 }, "model": { "tag": "RefinedNeuro/Turkcell-LLM-7b-v1:latest", "catalogSlug": "turkcell-llm-7b", "family": "llama", "parameterSize": "7.4B", "quantizationLevel": "Q4_K_M", "sizeBytes": 4461475966, "digest": "50f61cb255fad048b2e56684390e8397b69341f90644289d6182671ab197167d", "isTurkishModel": true }, "evidenceUrl": "https://www.runlocalai.co/benchmarks/evidence/local-ollama-rtx-5080-2026-05-28-turkish-sweep/turkcell-llm-7b/evidence.json", "command": "POST http://localhost:11434/api/generate model=RefinedNeuro/Turkcell-LLM-7b-v1:latest temperature=0 seed=42 num_ctx=2048 num_predict=512", "outputHash": "12c8d822940557b9a8381fcd5594d09e62b5ca1bcc036192447ba812eb8d688d", "sourceFileSha256": "897d20358d0a41bb640d93209207220980320a91618c1f3a18a382dbbe22e414", "sourceResultSha256": "6ffc8a6782bb9e94aa30524c89da9867d4caa5829a043f932e661c2d3b91d037", "artifactHashSha256": "1025398db148efe5695f432fbd7bed17af4d13623a5bcf73a929c39c101d8ab0", "logHashSha256": "1025398db148efe5695f432fbd7bed17af4d13623a5bcf73a929c39c101d8ab0", "logHashNote": "Structured Ollama API sweep hash. No separate verbose text logs were captured for this sweep.", "intakeVersion": "V39.0-ollama-tps-sweep", "capturedAt": "2026-05-28T05:13:10.545Z" }, "rawResult": { "model": "RefinedNeuro/Turkcell-LLM-7b-v1:latest", "model_slug": "turkcell-llm-7b", "is_turkish_model": true, "family": "llama", "parameter_size": "7.4B", "quantization_level": "Q4_K_M", "model_size_bytes": 4461475966, "digest": "50f61cb255fad048b2e56684390e8397b69341f90644289d6182671ab197167d", "scores_generation_tps": [ 146.35, 145.77, 144.04, 145.08, 144.99 ], "mean_generation_tps": 145.25, "median_generation_tps": 145.08, "min_generation_tps": 144.04, "max_generation_tps": 146.35, "runs": [ { "run": 1, "generated_tokens": 494, "prompt_tokens": 65, "eval_duration_ns": 3375520700, "prompt_eval_duration_ns": 20116000, "total_duration_ns": 3619655900, "wall_ms": 3623, "generation_tps": 146.35, "total_tps": 136.48, "done_reason": "stop", "response_sha256": "ac73f2fbd8e24088a967d02ab3534ba60916e20872c2be5a1e0fe9358d7854f8" }, { "run": 2, "generated_tokens": 500, "prompt_tokens": 65, "eval_duration_ns": 3429985100, "prompt_eval_duration_ns": 8722000, "total_duration_ns": 3691097200, "wall_ms": 3694, "generation_tps": 145.77, "total_tps": 135.46, "done_reason": "stop", "response_sha256": "f8ada0608c9ac3ac7b205e8300831ddf686a841afa34c3dc216085fb8d09035e" }, { "run": 3, "generated_tokens": 500, "prompt_tokens": 65, "eval_duration_ns": 3471178600, "prompt_eval_duration_ns": 9204000, "total_duration_ns": 3700920100, "wall_ms": 3704, "generation_tps": 144.04, "total_tps": 135.1, "done_reason": "stop", "response_sha256": "f8ada0608c9ac3ac7b205e8300831ddf686a841afa34c3dc216085fb8d09035e" }, { "run": 4, "generated_tokens": 500, "prompt_tokens": 65, "eval_duration_ns": 3446365800, "prompt_eval_duration_ns": 8610500, "total_duration_ns": 3679243900, "wall_ms": 3682, "generation_tps": 145.08, "total_tps": 135.9, "done_reason": "stop", "response_sha256": "f8ada0608c9ac3ac7b205e8300831ddf686a841afa34c3dc216085fb8d09035e" }, { "run": 5, "generated_tokens": 500, "prompt_tokens": 65, "eval_duration_ns": 3448557800, "prompt_eval_duration_ns": 8618600, "total_duration_ns": 3686043500, "wall_ms": 3690, "generation_tps": 144.99, "total_tps": 135.65, "done_reason": "stop", "response_sha256": "f8ada0608c9ac3ac7b205e8300831ddf686a841afa34c3dc216085fb8d09035e" } ] } }