{ "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/rn-tr-r2/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/RN_TR_R2: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": "38ef01097f7fe4f1436c15e06fece53a052c491deda10695d9a322f2c2121363", "logHashSha256": "38ef01097f7fe4f1436c15e06fece53a052c491deda10695d9a322f2c2121363", "logHashNote": "Structured Ollama API sweep hash. No separate verbose text logs were captured for this sweep.", "modelTag": "RefinedNeuro/RN_TR_R2:latest", "catalogSlug": "rn-tr-r2", "family": "llama", "parameterSize": "8.0B", "quantizationLevel": "Q4_K_M", "modelSizeBytes": 4920736342, "digest": "8eff3a89286d9e0d783df6582a22feaba32e9f5826c1b8fc89c1e1b3acb80858", "isTurkishModel": false, "scoresGenerationTps": [ 133.44, 133.15, 132.65, 133.58, 133.51 ], "statsGenerationTps": { "min": 132.65, "max": 133.58, "median": 133.44, "p5": 132.75, "p95": 133.566, "mean": 133.26600000000002, "stdev": 0.34097507240265945, "cvPct": 0.2558605138614946, "n": 5 }, "runsCaptured": 5, "runStatsJson": { "runs": [ { "runIndex": 1, "tokPerSec": 133.44, "totalTokPerSec": 118.73, "ttftMs": null, "prefillTokPerSec": 6900.320471259471, "promptTokens": 149, "responseTokens": 512, "promptEvalMs": 21.5932, "evalMs": 3836.867, "totalMs": 4312.31, "wallMs": 4315, "doneReason": "length", "responseHash": "032cd07799b32c13550f58523e0f6247b6138af6fa9b02c98cb24565d5222f6f" }, { "runIndex": 2, "tokPerSec": 133.15, "totalTokPerSec": 121.76, "ttftMs": null, "prefillTokPerSec": 16302.50445857085, "promptTokens": 149, "responseTokens": 512, "promptEvalMs": 9.1397, "evalMs": 3845.3096, "totalMs": 4204.8999, "wallMs": 4208, "doneReason": "length", "responseHash": "cb12da3e37df823438d2358dde0162653f1fe3612bcf390c05c2c846d7e2c192" }, { "runIndex": 3, "tokPerSec": 132.65, "totalTokPerSec": 121.19, "ttftMs": null, "prefillTokPerSec": 15365.576982571929, "promptTokens": 149, "responseTokens": 512, "promptEvalMs": 9.697, "evalMs": 3859.7119, "totalMs": 4224.8968, "wallMs": 4228, "doneReason": "length", "responseHash": "cb12da3e37df823438d2358dde0162653f1fe3612bcf390c05c2c846d7e2c192" }, { "runIndex": 4, "tokPerSec": 133.58, "totalTokPerSec": 120.06, "ttftMs": null, "prefillTokPerSec": 16329.661899282151, "promptTokens": 149, "responseTokens": 512, "promptEvalMs": 9.1245, "evalMs": 3832.8535, "totalMs": 4264.4117, "wallMs": 4268, "doneReason": "length", "responseHash": "cb12da3e37df823438d2358dde0162653f1fe3612bcf390c05c2c846d7e2c192" }, { "runIndex": 5, "tokPerSec": 133.51, "totalTokPerSec": 121.94, "ttftMs": null, "prefillTokPerSec": 15024.402048965434, "promptTokens": 149, "responseTokens": 512, "promptEvalMs": 9.9172, "evalMs": 3834.7976, "totalMs": 4198.6225, "wallMs": 4201, "doneReason": "length", "responseHash": "cb12da3e37df823438d2358dde0162653f1fe3612bcf390c05c2c846d7e2c192" } ], "coldStart": null, "steady": { "min": 132.65, "max": 133.58, "median": 133.44, "p5": 132.75, "p95": 133.566, "mean": 133.26600000000002, "stdev": 0.34097507240265945, "cvPct": 0.2558605138614946, "n": 5 }, "warm": { "min": 132.65, "max": 133.58, "median": 133.44, "p5": 132.75, "p95": 133.566, "mean": 133.26600000000002, "stdev": 0.34097507240265945, "cvPct": 0.2558605138614946, "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/RN_TR_R2:latest", "catalogSlug": "rn-tr-r2", "family": "llama", "parameterSize": "8.0B", "quantizationLevel": "Q4_K_M", "sizeBytes": 4920736342, "digest": "8eff3a89286d9e0d783df6582a22feaba32e9f5826c1b8fc89c1e1b3acb80858", "isTurkishModel": false }, "evidenceUrl": "https://www.runlocalai.co/benchmarks/evidence/local-ollama-rtx-5080-2026-05-28-turkish-sweep/rn-tr-r2/evidence.json", "command": "POST http://localhost:11434/api/generate model=RefinedNeuro/RN_TR_R2:latest temperature=0 seed=42 num_ctx=2048 num_predict=512", "outputHash": "12c8d822940557b9a8381fcd5594d09e62b5ca1bcc036192447ba812eb8d688d", "sourceFileSha256": "897d20358d0a41bb640d93209207220980320a91618c1f3a18a382dbbe22e414", "sourceResultSha256": "6ffc8a6782bb9e94aa30524c89da9867d4caa5829a043f932e661c2d3b91d037", "artifactHashSha256": "38ef01097f7fe4f1436c15e06fece53a052c491deda10695d9a322f2c2121363", "logHashSha256": "38ef01097f7fe4f1436c15e06fece53a052c491deda10695d9a322f2c2121363", "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/RN_TR_R2:latest", "model_slug": "rn-tr-r2", "is_turkish_model": false, "family": "llama", "parameter_size": "8.0B", "quantization_level": "Q4_K_M", "model_size_bytes": 4920736342, "digest": "8eff3a89286d9e0d783df6582a22feaba32e9f5826c1b8fc89c1e1b3acb80858", "scores_generation_tps": [ 133.44, 133.15, 132.65, 133.58, 133.51 ], "mean_generation_tps": 133.27, "median_generation_tps": 133.44, "min_generation_tps": 132.65, "max_generation_tps": 133.58, "runs": [ { "run": 1, "generated_tokens": 512, "prompt_tokens": 149, "eval_duration_ns": 3836867000, "prompt_eval_duration_ns": 21593200, "total_duration_ns": 4312310000, "wall_ms": 4315, "generation_tps": 133.44, "total_tps": 118.73, "done_reason": "length", "response_sha256": "032cd07799b32c13550f58523e0f6247b6138af6fa9b02c98cb24565d5222f6f" }, { "run": 2, "generated_tokens": 512, "prompt_tokens": 149, "eval_duration_ns": 3845309600, "prompt_eval_duration_ns": 9139700, "total_duration_ns": 4204899900, "wall_ms": 4208, "generation_tps": 133.15, "total_tps": 121.76, "done_reason": "length", "response_sha256": "cb12da3e37df823438d2358dde0162653f1fe3612bcf390c05c2c846d7e2c192" }, { "run": 3, "generated_tokens": 512, "prompt_tokens": 149, "eval_duration_ns": 3859711900, "prompt_eval_duration_ns": 9697000, "total_duration_ns": 4224896800, "wall_ms": 4228, "generation_tps": 132.65, "total_tps": 121.19, "done_reason": "length", "response_sha256": "cb12da3e37df823438d2358dde0162653f1fe3612bcf390c05c2c846d7e2c192" }, { "run": 4, "generated_tokens": 512, "prompt_tokens": 149, "eval_duration_ns": 3832853500, "prompt_eval_duration_ns": 9124500, "total_duration_ns": 4264411700, "wall_ms": 4268, "generation_tps": 133.58, "total_tps": 120.06, "done_reason": "length", "response_sha256": "cb12da3e37df823438d2358dde0162653f1fe3612bcf390c05c2c846d7e2c192" }, { "run": 5, "generated_tokens": 512, "prompt_tokens": 149, "eval_duration_ns": 3834797600, "prompt_eval_duration_ns": 9917200, "total_duration_ns": 4198622500, "wall_ms": 4201, "generation_tps": 133.51, "total_tps": 121.94, "done_reason": "length", "response_sha256": "cb12da3e37df823438d2358dde0162653f1fe3612bcf390c05c2c846d7e2c192" } ] } }