Fine-tuning Dataset Validator

Un errors ke liye conversation dataset jaanchiye jo training ko beech mein gira dete hain.

Training data browser mein padha jaata hai. Kuchh upload nahi hota aur panna band karne par kuchh bachta nahi.

File ko record-dar-record dekhiye

Dataset Validator ke baare mein

Training pehli hi kharab record par girta hai, aur aam taur par use tab dhoondhta hai jab aap ghar ja chuke hote hain. Pehle karne layak jaanchein maamuli aur tez hain: har record messages array wala object ho; har message mein role aur content ho; conversation wahin se shuru ho jahaan se honi chahiye; aur — yahi sabse zyada chhootti hai — assistant ke message par khatam ho, kyunki record jo sikhati hai wo wahi turn hai. Lagataar do user messages bhi dikhaye jaate hain, kyunki aam taur par uska matlab hai ki do records aapas mein jud gayin. Errors ke saath-saath aakaar ka vitaran us poonchh ko dikha deta hai jo samayegi nahi — isse pehle ki wahi rukne ki wajah bane.

Features

• Har note usi line par jahaan se wo uthi
• Har record padhi jaati hai ya nahi ye jaanchta hai aur aage badh jaata hai
• Roles gyaat samooh se milaye jaate hain, asaamanya dikhaye jaate hain
• Assistant ke message par khatam na hone wali record pakadta hai
• Ek hi paksh ki lagataar do turns pakadta hai
• Khaali content pakadta hai, jo jaanch paas kar leti hai aur kuchh sikhati nahi
• Aakaar ka vitaran: madhyika, 90waan percentile aur adhiktam
• Alag naam wale formats ke liye badalne yogya messages key

Dataset Validator kaise use karein

1. .jsonl dataset drop kijiye ya paste kijiye
2. Format alag ho to messages key badliye
3. Errors theek kijiye, phir warnings dekhiye
4. Aakaar ka vitaran apne context ki seema se milaiye

Example

Input

line 3: two user messages in a row

Output

Records: 3 · Errors: 0 · Warnings: 1

Aam taur par sanket ki do records galti se jud gayin.

Common errors aur troubleshooting

• Koi record assistant ke message par khatam nahi hoti. — Us record ke paas sikhane ko kuchh nahi hai: lakshya assistant ka aakhri turn hai. Isiliye ye warning nahi, error hai.
• messages array nahi mil raha. — messages key wahi rakhiye jo aapke format mein hai. conversation aur turns dono aam hain.
• Lagataar do user messages dikhaye ja rahe hain. — Aam taur par do records galti se jud gayi hain. Dohri turn jaan-boojhkar ho to ye warning hai, error nahi.
• Kuchh records bahut badi hain. — Percentiles poonchh dikha dete hain. Context seema se upar sab kuchh fail hoga, aur ye yahaan pata chalna training ke beech pata chalne se kahin sasta hai.

Aksar pooche jaane wale sawaal

Ye kaun-sa format chahta hai?

chat format: har line par ek JSON object, har ek mein role-content jodon wala messages array. Key ka naam badla ja sakta hai, par bhitar ka aakaar har jagah wahi hai.

Record assistant ke message par hi kyun khatam ho?

Kyunki record jo sikhati hai wo wahi turn hai. User ke message par khatam hoti conversation mein sawal hai aur jawab nahi, to seekhne ko kuchh hai hi nahi.

Kya token ki ginti sateek hai?

Nahi, ye aksharon par aadharit anumaan hain. Ye us apwaad ko pakadne ke liye hain jo samayega nahi, aur uske liye sateekta zaroori nahi: sateek sankhya provider ka tokenizer deta hai.

Saaf nateeje ka matlab data achha hai?

Matlab itna hai ki file ki banawat theek hai. Examples sahi cheez sikhate hain ya nahi, ye aapke data ka sawal hai aur uska jawab koi validator nahi deta.

Kya mera dataset upload hota hai?

Nahi. Training data aam taur par kisi team ki sabse sanvedanshil file hoti hai, aur ye sab browser mein chalta hai.

Related tools

• JSONL Viewer aur Validator — JSONL file line-dar-line padhiye, chhaantiye aur jaanchiye.
• Chat Transcript Converter — Messages array ko padhne layak transcript mein badliye, aur wapas.
• Batch Request Builder — Rows ki table ko alag pehchaan wale bulk job mein badliye.
• Text Chunker — Document ko asli seemayon par tootne wale overlap tukdon mein baantiye.
• Context Window Planner — Window ko prompt, tools, history aur jawab mein baantiye.
• Prompt Template Filler — Ek prompt ko rows ki table par failaiye aur kami dekhiye.

AI aur agents ke aur tools

• Skill Builder — SKILL.md likhiye, frontmatter saath-saath check hota hai.
• Skill Linter — SKILL.md paste kijiye aur dekhiye use kisi kaam se judne se kya rok raha hai.
• Subagent Builder — Subagent ki definition file banaiye, uske tools aur model samet.
• Slash Command Builder — Arguments aur tools list ke saath apna command banaiye.
• Project Instructions Builder — Aisi CLAUDE.md likhiye jo har session mein apni jagah kamaye.
• Plugin Manifest Builder — plugin.json banaiye aur uski apekshit folder banawat dekhiye.
• Hooks Builder — settings.json ke hooks banaiye aur kabhi na chalne wale matchers pakadiye.
• Permission Rule Tester — Allow, ask aur deny rules asli commands par aazmaiye.
• MCP Server Config — stdio, HTTP ya SSE ke liye MCP server entry banaiye.
• MCP Tool Schema Builder — Tool tay kijiye aur JSON Schema paiye, strict mode jaanch ke saath.
• Tool Schema Checker — Definition paste kijiye aur jaaniye wo kyun reject ho rahi hai.
• Prompt Template Filler — Ek prompt ko rows ki table par failaiye aur kami dekhiye.
• System Prompt Linter — Contradictions, bekaar instructions aur bina bhare placeholders dhoondhiye.
• Prompt Diff — Do versions shabd-dar-shabd milaiye, token antar ke saath.
• Few-shot Example Builder — Input-output jodon ko tag, JSON ya chat roop ke examples banaiye.
• XML Prompt Section Builder — Sections ko aise tags mein lapetiye jinse content bahar na nikle.
• JSONL Viewer aur Validator — JSONL file line-dar-line padhiye, chhaantiye aur jaanchiye.
• Chat Transcript Converter — Messages array ko padhne layak transcript mein badliye, aur wapas.
• SSE Stream Viewer — Kacchi stream se text, tool calls aur usage wapas paiye.
• Context Window Planner — Window ko prompt, tools, history aur jawab mein baantiye.
• Prompt Cache Planner — Blocks aise kram mein rakhiye ki cache hone wala prefix sabse lamba ho.
• Text Chunker — Document ko asli seemayon par tootne wale overlap tukdon mein baantiye.
• Agent Loop Cost Estimator — Turn-dar-turn kharch nikaaliye, dobara bheje history aur cache ke saath.
• Batch Request Builder — Rows ki table ko alag pehchaan wale bulk job mein badliye.

AI aur agents ke sabhi tools dekhein · Saare tools