Datasets:
audio audioduration (s) 24.2 37.1 | id stringlengths 23 23 | text stringlengths 428 683 | duration_s float64 24.2 37.1 | categories listlengths 1 1 | target_terms listlengths 3 6 | prompt_id stringlengths 20 20 | prompt_source stringclasses 1
value | mode stringclasses 1
value | edited bool 1
class | device stringclasses 1
value | environment stringclasses 1
value | sample_rate int64 48k 48k | recorded_at stringlengths 24 24 | microphone stringclasses 1
value | prompt_origin stringclasses 1
value | hebrew_pronunciation stringclasses 0
values | variant_groups listlengths 0 0 | variant_spelling stringclasses 0
values | speaker_id stringdate 8992-01-01 00:00:00 8992-01-01 00:00:00 | conditions listlengths 0 0 | imported_from stringclasses 0
values | original_text stringclasses 0
values | text_before_normalization stringclasses 1
value | text_he stringlengths 428 683 | noise_level stringclasses 0
values | background_speech_languages listlengths 0 0 | notes stringclasses 0
values | model_hypothesis stringclasses 0
values | asr_model stringclasses 0
values | speaker stringclasses 1
value | speaker_gender stringclasses 1
value | noise_flags listlengths 0 0 | clipped bool 1
class | task stringclasses 1
value | spoken_text stringlengths 428 683 | aside_text stringclasses 0
values | aside_position stringclasses 0
values | verbatim_text stringclasses 0
values | prompt_categories listlengths | source_prompt_id stringclasses 0
values | prompt_kind stringclasses 7
values | prompt_note stringclasses 0
values | term_language stringclasses 0
values | paragraph_text stringclasses 0
values | contributor stringclasses 1
value | bit_depth int64 16 24 | audio_format stringclasses 1
value | native_rate int64 48k 48k | capture_mode stringclasses 1
value | tts_eligible bool 1
class | tts_split stringclasses 0
values | transcript_status stringclasses 1
value | prompt_set stringclasses 0
values | topic stringclasses 0
values | style stringclasses 0
values | style_instruction stringclasses 0
values | written_form_uncertain bool 0
classes | session_id stringclasses 2
values | room_tone_id stringclasses 0
values | lead_trim_s float64 0 0 | peak_dbfs float64 -3.9 -1.5 | floor_dbfs float64 -44.9 -42.3 | speech_dbfs float64 -19.4 -10.1 | snr_db float64 23.5 32.6 | clipped_samples int64 0 0 | edge_head_dbfs float64 -90.3 -18.3 | edge_tail_dbfs float64 -39.3 -24.8 | browser_processing bool 1
class | room stringclasses 0
values | priority_word stringclasses 0
values | location stringclasses 0
values | location_name stringclasses 0
values | location_note stringclasses 0
values | latitude float64 | longitude float64 | location_accuracy_m float64 | text_full stringclasses 0
values | fallback_ok bool 0
classes | sentence_id stringlengths 12 12 | pronunciation stringclasses 1
value | extra stringclasses 2
values |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
20261106T000818Z_hd3a6a | Remember to sideload the APK onto my OnePlus tonight, because the Play Store version keeps crashing right after the login screen. Please set up the cluster config so everything follows GitOps, meaning no manual changes on the server, and every change goes through a pull request first. I want to move this whole server s... | 29.93 | [
"long-tech"
] | [
"APK",
"OnePlus",
"GitOps",
"devcontainer"
] | long-tech_315df0c714 | generated:composed-from-bank | bank | false | OnePlus Nord 3 5G | quiet-indoor | 48,000 | 2026-11-06T00:08:18.564Z | internal | generated | null | [] | null | 8992 | [] | null | null | null | Remember to sideload the APK onto my OnePlus tonight, because the Play Store version keeps crashing right after the login screen. Please set up the cluster config so everything follows GitOps, meaning no manual changes on the server, and every change goes through a pull request first. I want to move this whole server s... | null | [] | null | null | null | Daniel Rosehill | male | [] | false | long-tech | Remember to sideload the APK onto my OnePlus tonight, because the Play Store version keeps crashing right after the login screen. Please set up the cluster config so everything follows GitOps, meaning no manual changes on the server, and every change goes through a pull request first. I want to move this whole server s... | null | null | null | null | null | general_tech | null | null | null | Daniel Rosehill | 24 | flac | 48,000 | stt | true | null | script | null | null | null | null | null | 20261106T000236Z_k8qwxf | null | 0 | -3.9 | -42.7 | -18.7 | 24.1 | 0 | -35.5 | -33.3 | false | null | null | null | null | null | null | null | null | null | null | p_276a2aaf16 | [] | {"mic_link": "internal", "bt_codec": null, "mic_name": "CPH2493", "mic_usb_vendor_id": null, "mic_usb_product_id": null, "activity": "seated", "pronunciation_variants": [], "ui_note": null, "source_url": null, "check_verdict": null, "check_errors": null, "check_error_types": null} | |
20261106T000851Z_fg3vdd | Please check the TrueNAS box over SSH, list the ZFS pools and tell me which datasets are almost full before we move anything. I switched the internal event format over to Protobuf last week, so please regenerate your client stubs before you send me the next batch of test data. Check the Ceph cluster health on the stagi... | 30.479 | [
"long-tech"
] | [
"TrueNAS",
"Protobuf",
"Ceph"
] | long-tech_02b309d38f | generated:composed-from-bank | bank | false | OnePlus Nord 3 5G | quiet-indoor | 48,000 | 2026-11-06T00:08:51.141Z | internal | generated | null | [] | null | 8992 | [] | null | null | null | Please check the TrueNAS box over SSH, list the ZFS pools and tell me which datasets are almost full before we move anything. I switched the internal event format over to Protobuf last week, so please regenerate your client stubs before you send me the next batch of test data. Check the Ceph cluster health on the stagi... | null | [] | null | null | null | Daniel Rosehill | male | [] | false | long-tech | Please check the TrueNAS box over SSH, list the ZFS pools and tell me which datasets are almost full before we move anything. I switched the internal event format over to Protobuf last week, so please regenerate your client stubs before you send me the next batch of test data. Check the Ceph cluster health on the stagi... | null | null | null | null | null | homelab_network | null | null | null | Daniel Rosehill | 24 | flac | 48,000 | stt | true | null | script | null | null | null | null | null | 20261106T000236Z_k8qwxf | null | 0 | -1.5 | -42.3 | -18.5 | 23.8 | 0 | -36.5 | -33.9 | false | null | null | null | null | null | null | null | null | null | null | p_78b1008397 | [] | {"mic_link": "internal", "bt_codec": null, "mic_name": "CPH2493", "mic_usb_vendor_id": null, "mic_usb_product_id": null, "activity": "seated", "pronunciation_variants": [], "ui_note": null, "source_url": null, "check_verdict": null, "check_errors": null, "check_error_types": null} | |
20261106T000929Z_q6ce19 | Check the compose file for Navidrome and make sure the music folder is mounted read only, then restart the container and tell me if the library rescan finishes. I think we should switch that endpoint to GraphQL, because the mobile app keeps fetching way more fields than it actually needs. Check the DNS records for our ... | 31.69 | [
"long-tech"
] | [
"Navidrome",
"GraphQL",
"DKIM",
"UniFi",
"Jellyfin"
] | long-tech_4e8534013b | generated:composed-from-bank | bank | false | OnePlus Nord 3 5G | quiet-indoor | 48,000 | 2026-11-06T00:09:29.120Z | internal | generated | null | [] | null | 8992 | [] | null | null | null | Check the compose file for Navidrome and make sure the music folder is mounted read only, then restart the container and tell me if the library rescan finishes. I think we should switch that endpoint to GraphQL, because the mobile app keeps fetching way more fields than it actually needs. Check the DNS records for our ... | null | [] | null | null | null | Daniel Rosehill | male | [] | false | long-tech | Check the compose file for Navidrome and make sure the music folder is mounted read only, then restart the container and tell me if the library rescan finishes. I think we should switch that endpoint to GraphQL, because the mobile app keeps fetching way more fields than it actually needs. Check the DNS records for our ... | null | null | null | null | null | homelab_network | null | null | null | Daniel Rosehill | 24 | flac | 48,000 | stt | true | null | script | null | null | null | null | null | 20261106T000236Z_k8qwxf | null | 0 | -2.9 | -42.4 | -18.7 | 23.7 | 0 | -30.8 | -33.7 | false | null | null | null | null | null | null | null | null | null | null | p_bc6cdd8387 | [] | {"mic_link": "internal", "bt_codec": null, "mic_name": "CPH2493", "mic_usb_vendor_id": null, "mic_usb_product_id": null, "activity": "seated", "pronunciation_variants": [], "ui_note": null, "source_url": null, "check_verdict": null, "check_errors": null, "check_error_types": null} | |
20261106T001037Z_nje792 | Check which tokenizer the local model is using in the repo, and make sure the same one is loaded when we count tokens before sending each request. Please check how the tokenizer in this repo handles BPE merges, and add a small test that confirms rare words get split into sensible pieces. Please create a marimo notebook... | 24.255 | [
"long-tech"
] | [
"tokenizer",
"BPE",
"marimo"
] | long-tech_c1d96cc2ce | generated:composed-from-bank | bank | false | OnePlus Nord 3 5G | quiet-indoor | 48,000 | 2026-11-06T00:10:37.181Z | internal | generated | null | [] | null | 8992 | [] | null | null | null | Check which tokenizer the local model is using in the repo, and make sure the same one is loaded when we count tokens before sending each request. Please check how the tokenizer in this repo handles BPE merges, and add a small test that confirms rare words get split into sensible pieces. Please create a marimo notebook... | null | [] | null | null | null | Daniel Rosehill | male | [] | false | long-tech | Check which tokenizer the local model is using in the repo, and make sure the same one is loaded when we count tokens before sending each request. Please check how the tokenizer in this repo handles BPE merges, and add a small test that confirms rare words get split into sensible pieces. Please create a marimo notebook... | null | null | null | null | null | ai_ml | null | null | null | Daniel Rosehill | 24 | flac | 48,000 | stt | true | null | script | null | null | null | null | null | 20261106T000236Z_k8qwxf | null | 0 | -3.9 | -42.3 | -18.7 | 23.5 | 0 | -90.3 | -34.9 | false | null | null | null | null | null | null | null | null | null | null | p_b58751dcd6 | [] | {"mic_link": "internal", "bt_codec": null, "mic_name": "CPH2493", "mic_usb_vendor_id": null, "mic_usb_product_id": null, "activity": "seated", "pronunciation_variants": [], "ui_note": null, "source_url": null, "check_verdict": null, "check_errors": null, "check_error_types": null} | |
20261106T001143Z_wqxysx | Reminder to rebuild the blog tonight with the new SSG config, because the old build is still pulling in the draft pages. Please add PostHog event tracking to the signup flow, and make sure it only fires after the user accepts the cookie banner. Thinking of rebuilding my blog as Jamstack, static pages with a small API f... | 35.615 | [
"long-tech"
] | [
"SSG",
"PostHog",
"Jamstack",
"shadcn",
"AudioWorklet",
"Tailscale"
] | long-tech_c9b676cc4a | generated:composed-from-bank | bank | false | OnePlus Nord 3 5G | quiet-indoor | 48,000 | 2026-11-06T00:11:43.943Z | internal | generated | null | [] | null | 8992 | [] | null | null | null | Reminder to rebuild the blog tonight with the new SSG config, because the old build is still pulling in the draft pages. Please add PostHog event tracking to the signup flow, and make sure it only fires after the user accepts the cookie banner. Thinking of rebuilding my blog as Jamstack, static pages with a small API f... | null | [] | null | null | null | Daniel Rosehill | male | [] | false | long-tech | Reminder to rebuild the blog tonight with the new SSG config, because the old build is still pulling in the draft pages. Please add PostHog event tracking to the signup flow, and make sure it only fires after the user accepts the cookie banner. Thinking of rebuilding my blog as Jamstack, static pages with a small API f... | null | null | null | null | null | web | null | null | null | Daniel Rosehill | 24 | flac | 48,000 | stt | true | null | script | null | null | null | null | null | 20261106T000236Z_k8qwxf | null | 0 | -2.4 | -43.3 | -19 | 24.3 | 0 | -32.7 | -34.2 | false | null | null | null | null | null | null | null | null | null | null | p_fa570aef48 | [] | {"mic_link": "internal", "bt_codec": null, "mic_name": "CPH2493", "mic_usb_vendor_id": null, "mic_usb_product_id": null, "activity": "seated", "pronunciation_variants": [], "ui_note": null, "source_url": null, "check_verdict": null, "check_errors": null, "check_error_types": null} | |
20261106T001214Z_39xhwj | The squash and merge lost the co-author line, add it back in the message. Please look through the repository and tell me which dialogs still depend on Qt, then propose a plan to move them to the web interface. Remind me to read more about WebAssembly, because I want to try running that image filter directly in the brow... | 27.429 | [
"long-tech"
] | [
"squash and merge",
"Qt",
"WebAssembly",
"mise"
] | long-tech_b9f41a5f27 | generated:composed-from-bank | bank | false | OnePlus Nord 3 5G | quiet-indoor | 48,000 | 2026-11-06T00:12:14.984Z | internal | generated | null | [] | null | 8992 | [] | null | null | null | The squash and merge lost the co-author line, add it back in the message. Please look through the repository and tell me which dialogs still depend on Qt, then propose a plan to move them to the web interface. Remind me to read more about WebAssembly, because I want to try running that image filter directly in the brow... | null | [] | null | null | null | Daniel Rosehill | male | [] | false | long-tech | The squash and merge lost the co-author line, add it back in the message. Please look through the repository and tell me which dialogs still depend on Qt, then propose a plan to move them to the web interface. Remind me to read more about WebAssembly, because I want to try running that image filter directly in the brow... | null | null | null | null | null | dev_tools | null | null | null | Daniel Rosehill | 24 | flac | 48,000 | stt | true | null | script | null | null | null | null | null | 20261106T000236Z_k8qwxf | null | 0 | -2.8 | -43.6 | -18.5 | 25.1 | 0 | -32.8 | -32 | false | null | null | null | null | null | null | null | null | null | null | p_0b33845964 | [] | {"mic_link": "internal", "bt_codec": null, "mic_name": "CPH2493", "mic_usb_vendor_id": null, "mic_usb_product_id": null, "activity": "seated", "pronunciation_variants": [], "ui_note": null, "source_url": null, "check_verdict": null, "check_errors": null, "check_error_types": null} | |
20261106T001256Z_921b0r | The paper got rejected, and you can read the whole back and forth on OpenReview. We tested a new STT model last week, and it handles my accent much better than the old one, especially with Hebrew names. Write a training script using TRL to fine-tune the small model on our support conversations, and start with a short t... | 36.066 | [
"long-tech"
] | [
"OpenReview",
"STT",
"TRL",
"LoRA",
"webhooks",
"PyArrow"
] | long-tech_4a1ecb2abb | generated:composed-from-bank | bank | false | OnePlus Nord 3 5G | quiet-indoor | 48,000 | 2026-11-06T00:12:56.706Z | internal | generated | null | [] | null | 8992 | [] | null | null | null | The paper got rejected, and you can read the whole back and forth on OpenReview. We tested a new STT model last week, and it handles my accent much better than the old one, especially with Hebrew names. Write a training script using TRL to fine-tune the small model on our support conversations, and start with a short t... | null | [] | null | null | null | Daniel Rosehill | male | [] | false | long-tech | The paper got rejected, and you can read the whole back and forth on OpenReview. We tested a new STT model last week, and it handles my accent much better than the old one, especially with Hebrew names. Write a training script using TRL to fine-tune the small model on our support conversations, and start with a short t... | null | null | null | null | null | ai_ml | null | null | null | Daniel Rosehill | 24 | flac | 48,000 | stt | true | null | script | null | null | null | null | null | 20261106T000236Z_k8qwxf | null | 0 | -2.4 | -42.4 | -17.6 | 24.8 | 0 | -36 | -32.4 | false | null | null | null | null | null | null | null | null | null | null | p_c938ca3093 | [] | {"mic_link": "internal", "bt_codec": null, "mic_name": "CPH2493", "mic_usb_vendor_id": null, "mic_usb_product_id": null, "activity": "seated", "pronunciation_variants": [], "ui_note": null, "source_url": null, "check_verdict": null, "check_errors": null, "check_error_types": null} | |
20261106T001335Z_bw39x4 | Don't try to download all of Common Crawl, just sample a few files from the latest snapshot to check the Hebrew share. Please check whether ESPnet is already installed in the virtual environment on the GPU server, and if it isn't, set it up and run the smallest recipe. I finally got the MCP server connected to my notes... | 37.05 | [
"long-tech"
] | [
"Common Crawl",
"ESPnet",
"MCP",
"LAION",
"subagent",
"Google Dataset Search"
] | long-tech_c5db268720 | generated:composed-from-bank | bank | false | OnePlus Nord 3 5G | quiet-indoor | 48,000 | 2026-11-06T00:13:35.836Z | internal | generated | null | [] | null | 8992 | [] | null | null | null | Don't try to download all of Common Crawl, just sample a few files from the latest snapshot to check the Hebrew share. Please check whether ESPnet is already installed in the virtual environment on the GPU server, and if it isn't, set it up and run the smallest recipe. I finally got the MCP server connected to my notes... | null | [] | null | null | null | Daniel Rosehill | male | [] | false | long-tech | Don't try to download all of Common Crawl, just sample a few files from the latest snapshot to check the Hebrew share. Please check whether ESPnet is already installed in the virtual environment on the GPU server, and if it isn't, set it up and run the smallest recipe. I finally got the MCP server connected to my notes... | null | null | null | null | null | ai_ml | null | null | null | Daniel Rosehill | 24 | flac | 48,000 | stt | true | null | script | null | null | null | null | null | 20261106T000236Z_k8qwxf | null | 0 | -2.5 | -43 | -18.6 | 24.4 | 0 | -37.6 | -33.2 | false | null | null | null | null | null | null | null | null | null | null | p_3106a8f73b | [] | {"mic_link": "internal", "bt_codec": null, "mic_name": "CPH2493", "mic_usb_vendor_id": null, "mic_usb_product_id": null, "activity": "seated", "pronunciation_variants": [], "ui_note": null, "source_url": null, "check_verdict": null, "check_errors": null, "check_error_types": null} | |
20261106T001405Z_ric095 | Please spin up Langflow in a Docker container on the staging server, expose it only on the local network, and write the steps in the README. After the last compaction you forgot the branch name, so check git status again. Even in auto mode, stop and ask me before you delete anything on the server. Don't write a new mem... | 24.165 | [
"long-tech"
] | [
"Langflow",
"compaction",
"auto mode",
"memory file",
"hooks"
] | long-tech_3fc1b89f41 | generated:composed-from-bank | bank | false | OnePlus Nord 3 5G | quiet-indoor | 48,000 | 2026-11-06T00:14:05.679Z | internal | generated | null | [] | null | 8992 | [] | null | null | Please spin up Langflow in a Docker container on the staging server, expose it only on the local network, and write the steps in the readme. After the last compaction you forgot the branch name, so check git status again. Even in auto mode, stop and ask me before you delete anything on the server. Don't write a new mem... | Please spin up Langflow in a Docker container on the staging server, expose it only on the local network, and write the steps in the readme. After the last compaction you forgot the branch name, so check git status again. Even in auto mode, stop and ask me before you delete anything on the server. Don't write a new mem... | null | [] | null | null | null | Daniel Rosehill | male | [] | false | long-tech | Please spin up Langflow in a Docker container on the staging server, expose it only on the local network, and write the steps in the readme. After the last compaction you forgot the branch name, so check git status again. Even in auto mode, stop and ask me before you delete anything on the server. Don't write a new mem... | null | null | null | null | null | ai_agents | null | null | null | Daniel Rosehill | 24 | flac | 48,000 | stt | true | null | script | null | null | null | null | null | 20261106T000236Z_k8qwxf | null | 0 | -2.7 | -44.9 | -18.7 | 26.3 | 0 | -33.2 | -36.5 | false | null | null | null | null | null | null | null | null | null | null | p_768889011c | [] | {"mic_link": "internal", "bt_codec": null, "mic_name": "CPH2493", "mic_usb_vendor_id": null, "mic_usb_product_id": null, "activity": "seated", "pronunciation_variants": [], "ui_note": null, "source_url": null, "check_verdict": null, "check_errors": null, "check_error_types": null} | |
20261106T001438Z_2tdskb | Add a small script that sends the test prompts to Together AI instead of running them locally, and read the API key from the environment file. Please update the nightly backup script so it uploads the compressed database dump to our Backblaze bucket, and then deletes local copies older than a week. That image is on Qua... | 28.607 | [
"long-tech"
] | [
"Together AI",
"Backblaze",
"Quay",
"Pulumi"
] | long-tech_1c3ab029e7 | generated:composed-from-bank | bank | false | OnePlus Nord 3 5G | quiet-indoor | 48,000 | 2026-11-06T00:14:38.816Z | internal | generated | null | [] | null | 8992 | [] | null | null | null | Add a small script that sends the test prompts to Together AI instead of running them locally, and read the API key from the environment file. Please update the nightly backup script so it uploads the compressed database dump to our Backblaze bucket, and then deletes local copies older than a week. That image is on Qua... | null | [] | null | null | null | Daniel Rosehill | male | [] | false | long-tech | Add a small script that sends the test prompts to Together AI instead of running them locally, and read the API key from the environment file. Please update the nightly backup script so it uploads the compressed database dump to our Backblaze bucket, and then deletes local copies older than a week. That image is on Qua... | null | null | null | null | null | cloud_devops | null | null | null | Daniel Rosehill | 24 | flac | 48,000 | stt | true | null | script | null | null | null | null | null | 20261106T000236Z_k8qwxf | null | 0 | -3.3 | -44.2 | -19.4 | 24.9 | 0 | -35.3 | -39.3 | false | null | null | null | null | null | null | null | null | null | null | p_0576b3a039 | [] | {"mic_link": "internal", "bt_codec": null, "mic_name": "CPH2493", "mic_usb_vendor_id": null, "mic_usb_product_id": null, "activity": "seated", "pronunciation_variants": [], "ui_note": null, "source_url": null, "check_verdict": null, "check_errors": null, "check_error_types": null} | |
20261106T001514Z_hdzftd | Please go through the open Dependabot pull requests in this repo, merge the patch updates if the tests pass, and leave the major version bumps for me to review. I moved all our private repositories over to Gitea on the home server last night, so please update your remotes and let me know if anything breaks. Sitting by ... | 34.247 | [
"long-tech"
] | [
"Dependabot",
"Gitea",
"Quay",
"Portainer",
"Terraform"
] | long-tech_7911acf5cf | generated:composed-from-bank | bank | false | OnePlus Nord 3 5G | quiet-indoor | 48,000 | 2026-11-06T00:15:14.848Z | internal | generated | null | [] | null | 8992 | [] | null | null | null | Please go through the open Dependabot pull requests in this repo, merge the patch updates if the tests pass, and leave the major version bumps for me to review. I moved all our private repositories over to Gitea on the home server last night, so please update your remotes and let me know if anything breaks. Sitting by ... | null | [] | null | null | null | Daniel Rosehill | male | [] | false | long-tech | Please go through the open Dependabot pull requests in this repo, merge the patch updates if the tests pass, and leave the major version bumps for me to review. I moved all our private repositories over to Gitea on the home server last night, so please update your remotes and let me know if anything breaks. Sitting by ... | null | null | null | null | null | cloud_devops | null | null | null | Daniel Rosehill | 24 | flac | 48,000 | stt | true | null | script | null | null | null | null | null | 20261106T000236Z_k8qwxf | null | 0 | -1.5 | -44.4 | -17.7 | 26.7 | 0 | -41.2 | -37.1 | false | null | null | null | null | null | null | null | null | null | null | p_b9e56e7493 | [] | {"mic_link": "internal", "bt_codec": null, "mic_name": "CPH2493", "mic_usb_vendor_id": null, "mic_usb_product_id": null, "activity": "seated", "pronunciation_variants": [], "ui_note": null, "source_url": null, "check_verdict": null, "check_errors": null, "check_error_types": null} | |
20261120T064611Z_mulvwm | Please run mypy across the whole repo and fix the type errors, then check that the LXC container on the server still starts cleanly. Replace all the old emoji icons in the dashboard with Lucide icons, keep the sizes consistent, and don't touch the dark mode styling. Free demos on Hugging Face Spaces go to sleep after a... | 30.167 | [
"long-tech"
] | [
"mypy",
"LXC",
"Lucide",
"Hugging Face Spaces",
"gitignore",
"force push"
] | long-tech_d85f6e3293 | generated:composed-from-bank | bank | false | OnePlus Nord 3 5G | quiet-indoor | 48,000 | 2026-11-20T06:46:11.965Z | internal | generated | null | [] | null | 8992 | [] | null | null | null | Please run mypy across the whole repo and fix the type errors, then check that the LXC container on the server still starts cleanly. Replace all the old emoji icons in the dashboard with Lucide icons, keep the sizes consistent, and don't touch the dark mode styling. Free demos on Hugging Face Spaces go to sleep after a... | null | [] | null | null | null | Daniel Rosehill | male | [] | false | long-tech | Please run mypy across the whole repo and fix the type errors, then check that the LXC container on the server still starts cleanly. Replace all the old emoji icons in the dashboard with Lucide icons, keep the sizes consistent, and don't touch the dark mode styling. Free demos on Hugging Face Spaces go to sleep after a... | null | null | null | null | null | dev_tools | null | null | null | Daniel Rosehill | 16 | flac | 48,000 | stt | true | null | script | null | null | null | null | null | 20261120T062949Z_lsynu9 | null | 0 | -2.9 | -42.6 | -10.1 | 32.6 | 0 | -18.3 | -24.8 | false | null | null | null | null | null | null | null | null | null | null | p_a19d947417 | [] | {"mic_link":null,"bt_codec":null,"mic_name":"CPH2493","mic_usb_vendor_id":null,"mic_usb_product_id":null,"mic_type":"builtin_mic","mic_chain":null,"activity":"seated","background_speech_kinds":[],"pronunciation_variants":[],"ui_note":null,"source_url":null,"check_verdict":null,"check_errors":null,"check_error_types":nu... |
Long Tech Sentences
Audio dataset, split-out. Rows copied from the master, danielrosehill/Read-Aloud-Voice-Sentences, so one task can be browsed alone. Training reads the master, not this copy. Every audio dataset as one tree, with hours: the audio index. How every dataset fits together: the dataset index.
One speaker reading dictation-length tech passages, about 25 to 55 seconds in one take. Each passage joins 4 to 7 generated tech sentences from the same domain (scripts/build_long_passages.py), to match the length of real dictation. No Hebrew. Experimental.
Website: hebrish.app: get the Hebrish keyboard for Android and the desktop apps, see the results, and find every model, dataset and demo in one place.
Two tabs. sentences lists every ground-truth sentence with a takes count, so
takes = 0 shows what has not been recorded. takes has one row per recording with its audio, joined on sentence_id; a
sentence can have several takes. load_dataset("Hebrish/Long-Tech-Sentences", "sentences") reads the text; the default config reads the takes.
Takes come from danielrosehill/Read-Aloud-Voice-Sentences, the full dataset, with the same ids and
train/test split, by scripts/sync_subsets.py in the project repository. Do not edit this dataset directly. New takes append and
only the sentence counts are rewritten.
Speaker: Daniel Rosehill, the dataset's author, who was born in Ireland and speaks English with an Irish accent, and who lives in Jerusalem. Every recording is his voice. Deliberate variation: he sometimes changes his speaking speed on purpose (fast and rushed, or slow) and sometimes says a word the way people commonly mispronounce it, to make the models robust to real speech. The transcript is always the sentence as written, so a take whose speed or pronunciation sounds off against its card is usually intentional, not a recording error.
All audio was recorded on a OnePlus Nord 3 5G phone using its internal microphone, with browser echo cancellation, noise suppression and auto gain turned off.
Layout
sentences/train.parquet one row per sentence: `id`, `text`, `takes`, `prompt_ids`, `served` and the sentence columns
data/train-*.parquet one row per take: `audio` (the FLAC or WAV file, embedded) and every column
meta/train.jsonl index: the same rows without audio, plus `shard`, `audio_path` and the audio's byte range
meta/schema.json column types
data/test-*.parquet, meta/test.jsonl held-out set, never trained on
Audio is 48 kHz mono 24-bit FLAC for rows from 2026-10-22 and 16 kHz mono 16-bit WAV before.
The test split is term-disjoint: a held-out term never appears in train.
load_dataset("Hebrish/Long-Tech-Sentences") works directly.
Columns
A take's sentence_id is its row's id in sentences. A sentence with served: false is no longer offered by the app; it is kept because a take used it. The columns are the same as in the main dataset: text (ground truth, punctuated and cased like
Parakeet v2 output), text_he (the same transcript with its Hebrew terms in Hebrew script, for
Hebrew-capable models), categories, target_terms, prompt_source, prompt_origin (manual / dataset / generated), edited, duration_s,
environment, conditions, device, speaker, speaker_gender, contributor, hebrew_pronunciation, variant_groups, variant_spelling and recorded_at. The main dataset's card describes them.
Licence
CC-BY-4.0, with sentence-text attribution per prompt_source as described in the
main dataset's licence section.
- Downloads last month
- 112