lql-cli 0.2.1__tar.gz → 0.4.0__tar.gz

{lql_cli-0.2.1 → lql_cli-0.4.0}/.gitignore

@@ -13,3 +13,5 @@ venv/
 
 # local config
 ~/.lql/
+
+.env

{lql_cli-0.2.1 → lql_cli-0.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: lql-cli
-Version: 0.2.1
+Version: 0.4.0
 Summary: lql — CLI for the Liquid DataViewer platform
 Project-URL: Homepage, https://github.com/Liquid4All/lql
 Author: Liquid AI
@@ -9,6 +9,7 @@ Requires-Python: >=3.9
 Requires-Dist: httpx>=0.27
 Requires-Dist: huggingface-hub>=0.24
 Requires-Dist: rich>=13.0
+Requires-Dist: textual>=0.50
 Requires-Dist: typer>=0.12
 Description-Content-Type: text/markdown
 
@@ -58,7 +59,9 @@ cd lql
 uv tool install .     # or: pip install -e .
 ```
 
-Update to the latest version with `uv tool upgrade lql-cli` (or `pipx upgrade lql-cli` / `pip install -U lql-cli`).
+Update to the latest version with `lql update --run`, which detects how lql was
+installed (uv tool / pipx / pip) and runs the matching upgrade. Run `lql update`
+(no flag) to just print the command without executing it.
 
 ## Authentication
 
@@ -137,6 +140,50 @@ lql datasets upload <file> --workspace <id> --name <repo-name> [--split <s>]
 
 `datasets upload` requires `LQL_HF_TOKEN`.
 
+### Preview
+
+View dataset samples in the terminal — a Textual TUI with a chat-style layout
+(user turns on the right, system/assistant/tool on the left). It renders the
+same formats as the DataViewer web UI: OpenAI `{role, content}`, structured/
+multimodal content (text/image/audio), ShareGPT `{from, value}`, native OpenAI
+`tool_calls`, plus `<think>` reasoning blocks, `<|tool_call_start|>…<|tool_call_end|>`
+/ Python / XML / JSON tool calls, tool results, tool-definition tables, and code.
+
+Works on a local `.jsonl`/`.json` file, a platform dataset ID, or — with `--hf`
+— a HuggingFace repo. No browser, and nothing to forward over SSH — it's just
+the terminal.
+
+```
+lql preview <file.jsonl|file.json>     Local file: each line/object is a row
+lql preview <dataset-id>               Platform dataset (fetched & paged lazily)
+lql preview <org/name> --hf            HuggingFace repo: sync to DataViewer, then view
+lql preview <src> -c <field>           Force field(s) as conversations (repeatable)
+lql preview <src> -n <N>               Page size when paging a platform dataset
+lql preview <src> --offset N           Start at row index N
+lql preview <src> --title "<title>"    Title shown in the viewer header
+```
+
+**HuggingFace datasets (`--hf`).** `lql preview org/name --hf` syncs the repo
+into a DataViewer workspace, then opens it. You pick the target workspace from
+an interactive list (or pass `--workspace <id>`; `--split` defaults to `train`).
+Already-synced repos are reused — no duplicate, instant re-open.
+
+```
+lql preview tatsu-lab/alpaca --hf
+lql preview org/name --hf --split validation --workspace <id>
+```
+
+**Navigation** — two modes, toggle with `m` / `Tab`:
+
+- **pager** (default): one sample at a time · `←/→` or `n`/`b` switch samples · `↑/↓`/`j`/`k`/PgUp-Dn scroll
+- **scroll**: all samples in one buffer · `n`/`b` jump between samples · arrows scroll
+- `q` quits
+
+```
+lql preview examples/agent-traces.jsonl     # 20-sample file of agent-trace/tool-use formats
+lql preview <dataset-id>                     # browse a platform dataset, paged on demand
+```
+
 ### Evals (dataset analysis)
 
 Eval datasets are evaluation-run output — each row a sample with a model
@@ -284,6 +331,13 @@ lql instructions                                 Print the full agent reference
                                                  flags, examples, and workflows) in one read
 ```
 
+### Update
+
+```
+lql update                                       Detect install method (uv/pipx/pip), print upgrade cmd
+lql update --run                                 Detect and run the upgrade in place
+```
+
 ## Global flags
 
 All commands accept:

{lql_cli-0.2.1 → lql_cli-0.4.0}/README.md

@@ -44,7 +44,9 @@ cd lql
 uv tool install .     # or: pip install -e .
 ```
 
-Update to the latest version with `uv tool upgrade lql-cli` (or `pipx upgrade lql-cli` / `pip install -U lql-cli`).
+Update to the latest version with `lql update --run`, which detects how lql was
+installed (uv tool / pipx / pip) and runs the matching upgrade. Run `lql update`
+(no flag) to just print the command without executing it.
 
 ## Authentication
 
@@ -123,6 +125,50 @@ lql datasets upload <file> --workspace <id> --name <repo-name> [--split <s>]
 
 `datasets upload` requires `LQL_HF_TOKEN`.
 
+### Preview
+
+View dataset samples in the terminal — a Textual TUI with a chat-style layout
+(user turns on the right, system/assistant/tool on the left). It renders the
+same formats as the DataViewer web UI: OpenAI `{role, content}`, structured/
+multimodal content (text/image/audio), ShareGPT `{from, value}`, native OpenAI
+`tool_calls`, plus `<think>` reasoning blocks, `<|tool_call_start|>…<|tool_call_end|>`
+/ Python / XML / JSON tool calls, tool results, tool-definition tables, and code.
+
+Works on a local `.jsonl`/`.json` file, a platform dataset ID, or — with `--hf`
+— a HuggingFace repo. No browser, and nothing to forward over SSH — it's just
+the terminal.
+
+```
+lql preview <file.jsonl|file.json>     Local file: each line/object is a row
+lql preview <dataset-id>               Platform dataset (fetched & paged lazily)
+lql preview <org/name> --hf            HuggingFace repo: sync to DataViewer, then view
+lql preview <src> -c <field>           Force field(s) as conversations (repeatable)
+lql preview <src> -n <N>               Page size when paging a platform dataset
+lql preview <src> --offset N           Start at row index N
+lql preview <src> --title "<title>"    Title shown in the viewer header
+```
+
+**HuggingFace datasets (`--hf`).** `lql preview org/name --hf` syncs the repo
+into a DataViewer workspace, then opens it. You pick the target workspace from
+an interactive list (or pass `--workspace <id>`; `--split` defaults to `train`).
+Already-synced repos are reused — no duplicate, instant re-open.
+
+```
+lql preview tatsu-lab/alpaca --hf
+lql preview org/name --hf --split validation --workspace <id>
+```
+
+**Navigation** — two modes, toggle with `m` / `Tab`:
+
+- **pager** (default): one sample at a time · `←/→` or `n`/`b` switch samples · `↑/↓`/`j`/`k`/PgUp-Dn scroll
+- **scroll**: all samples in one buffer · `n`/`b` jump between samples · arrows scroll
+- `q` quits
+
+```
+lql preview examples/agent-traces.jsonl     # 20-sample file of agent-trace/tool-use formats
+lql preview <dataset-id>                     # browse a platform dataset, paged on demand
+```
+
 ### Evals (dataset analysis)
 
 Eval datasets are evaluation-run output — each row a sample with a model
@@ -270,6 +316,13 @@ lql instructions                                 Print the full agent reference
                                                  flags, examples, and workflows) in one read
 ```
 
+### Update
+
+```
+lql update                                       Detect install method (uv/pipx/pip), print upgrade cmd
+lql update --run                                 Detect and run the upgrade in place
+```
+
 ## Global flags
 
 All commands accept:

lql_cli-0.4.0/examples/agent-traces.jsonl

@@ -0,0 +1,20 @@
+{"id": "chat-001", "source": "openai-chat", "messages": [{"role": "system", "content": "You are a concise assistant."}, {"role": "user", "content": "What's the capital of France?"}, {"role": "assistant", "content": "The capital of France is **Paris**."}], "score": 1.0}
+{"id": "bfcl-002", "tool_format": "Python", "messages": [{"role": "system", "content": "You are an expert function-calling assistant. Here is a list of functions in JSON format that you can invoke:\n[{\"name\": \"get_weather\", \"description\": \"Get current weather for a city.\", \"parameters\": {\"type\":\"object\",\"properties\":{\"city\":{\"type\":\"string\",\"description\":\"City name\"},\"unit\":{\"type\":\"string\",\"description\":\"celsius or fahrenheit\"}},\"required\":[\"city\"]}}]"}, {"role": "user", "content": "What's the weather in Tokyo right now?"}, {"role": "assistant", "content": "<think>\nThe user wants current weather in Tokyo. I'll call get_weather with city=Tokyo.\n</think>\n<|tool_call_start|>[get_weather(city='Tokyo', unit='celsius')]<|tool_call_end|>"}, {"role": "tool", "content": "[{\"name\": \"get_weather\", \"result\": {\"city\": \"Tokyo\", \"temp_c\": 22, \"condition\": \"Partly cloudy\"}}]"}, {"role": "assistant", "content": "It's currently **22\u00b0C** and partly cloudy in Tokyo."}]}
+{"id": "sharegpt-003", "conversations": [{"from": "system", "value": "You are a helpful coding tutor."}, {"from": "human", "value": "How do I reverse a list in Python?"}, {"from": "gpt", "value": "Use slicing: `my_list[::-1]` or `list(reversed(my_list))`."}]}
+{"id": "reason-004", "messages": [{"role": "user", "content": "If a train travels 60 km in 45 minutes, what is its speed in km/h?"}, {"role": "assistant", "content": "<think>\n45 minutes = 0.75 hours. Speed = distance / time = 60 / 0.75 = 80.\n</think>\nThe train's speed is **80 km/h**."}]}
+{"id": "openai-native-005", "messages": [{"role": "user", "content": "Book a table for 2 at 7pm."}, {"role": "assistant", "content": null, "tool_calls": [{"id": "call_1", "type": "function", "function": {"name": "book_table", "arguments": "{\"party_size\": 2, \"time\": \"19:00\"}"}}]}, {"role": "tool", "tool_call_id": "call_1", "name": "book_table", "content": "{\"status\": \"confirmed\", \"reservation_id\": \"R-8842\"}"}, {"role": "assistant", "content": "Done! Your table for 2 at 7:00pm is confirmed (reservation R-8842)."}]}
+{"id": "parallel-006", "messages": [{"role": "user", "content": "Compare the weather in Paris and Berlin."}, {"role": "assistant", "content": "<|tool_call_start|>[get_weather(city='Paris'), get_weather(city='Berlin')]<|tool_call_end|>"}, {"role": "tool", "content": "[{\"name\": \"get_weather\", \"result\": {\"city\": \"Paris\", \"temp_c\": 18}}, {\"name\": \"get_weather\", \"result\": {\"city\": \"Berlin\", \"temp_c\": 15}}]"}, {"role": "assistant", "content": "Paris is 18\u00b0C and Berlin is 15\u00b0C \u2014 Paris is 3 degrees warmer."}]}
+{"id": "toolresult-007", "messages": [{"role": "user", "content": "Search for papers on RLHF."}, {"role": "assistant", "content": "<|tool_call_start|>[search(query='RLHF papers', limit=2)]<|tool_call_end|>"}, {"role": "tool", "content": "[{\"name\": \"search\", \"result\": [{\"title\": \"Deep RL from Human Preferences\", \"year\": 2017}, {\"title\": \"InstructGPT\", \"year\": 2022}]}]"}]}
+{"id": "xml-008", "messages": [{"role": "user", "content": "Convert 100 USD to EUR."}, {"role": "assistant", "content": "Let me convert that.\n<|tool_call_start|><function_call><name>convert_currency</name><arguments><amount>100</amount><from>USD</from><to>EUR</to></arguments></function_call><|tool_call_end|>"}, {"role": "tool", "content": "[{\"name\": \"convert_currency\", \"result\": {\"converted\": 92.4, \"rate\": 0.924}}]"}]}
+{"id": "multimodal-009", "messages": [{"role": "user", "content": [{"type": "text", "text": "What's in this image?"}, {"type": "image_url", "image_url": {"url": "https://example.com/cat.jpg"}}]}, {"role": "assistant", "content": "It appears to be a cat sitting on a windowsill."}]}
+{"id": "agenttrove-010", "agent": "terminus", "model": "gpt-4", "conversations": [{"role": "user", "content": "Solve the task. Respond as JSON with analysis, plan, commands."}, {"role": "assistant", "content": "{\n  \"analysis\": \"Need to create a script that checks balanced parentheses.\",\n  \"plan\": \"Write a Python script then test it.\",\n  \"commands\": [{\"keystrokes\": \"python3 solve.py\\n\", \"duration\": 0.5}]\n}"}]}
+{"id": "code-011", "messages": [{"role": "user", "content": "Write a Python function to compute factorial."}, {"role": "assistant", "content": "Here's a recursive implementation:\n\n```python\ndef factorial(n):\n    if n <= 1:\n        return 1\n    return n * factorial(n - 1)\n```\n\nIt runs in O(n) time."}]}
+{"id": "policy-012", "domain": "telecom", "messages": [{"role": "system", "content": "<instructions>\nYou are a customer service agent. Follow the policy.\n</instructions>\n<policy>\n# Telecom Policy\n- Verify identity before account changes.\n- Never disclose another customer's data.\n</policy>"}, {"role": "user", "content": "I want to change my plan."}, {"role": "assistant", "content": "I can help with that. First, can you verify your account by confirming your phone number?"}]}
+{"id": "memory-013", "domain": "volunteering", "messages": [{"role": "user", "content": "Remember my shift is Saturday 9am-12pm."}, {"role": "assistant", "content": "<think>\nStore the shift schedule in archival memory.\n</think>\n<|tool_call_start|>[archival_memory_add(key='shift', value='Saturday 09:00-12:00')]<|tool_call_end|>"}, {"role": "tool", "content": "[{\"name\": \"archival_memory_add\", \"result\": {\"status\": \"success\", \"total_entries\": 1}}]"}, {"role": "assistant", "content": "Got it \u2014 I've saved your Saturday 9am-12pm shift."}]}
+{"id": "tau2-014", "reward": 1.0, "domain": "telecom", "messages": [{"role": "assistant", "content": "Hi! How can I help you today?"}, {"role": "user", "content": "My mobile data is really slow."}, {"role": "assistant", "content": "<|tool_call_start|>[transfer_to_human_agents(summary='Customer reports slow mobile data.')]<|tool_call_end|>"}, {"role": "tool", "content": "Transfer successful"}, {"role": "assistant", "content": "I've transferred you to a human agent who can troubleshoot your connection."}]}
+{"id": "websearch-015", "num_hops": 2, "messages": [{"role": "system", "content": "You are an expert at composing function calls to answer multi-hop questions."}, {"role": "user", "content": "What language is spoken in the birthplace of the inventor of the telephone?"}, {"role": "assistant", "content": "<think>\nFirst find the inventor of the telephone, then their birthplace, then the language.\n</think>\n<|tool_call_start|>[search_engine_query(keywords='inventor of the telephone birthplace')]<|tool_call_end|>"}, {"role": "tool", "content": "[{\"name\": \"search_engine_query\", \"result\": [{\"url\": \"https://en.wikipedia.org/wiki/Alexander_Graham_Bell\", \"title\": \"Alexander Graham Bell - born Edinburgh, Scotland\"}]}]"}, {"role": "assistant", "content": "Bell was born in Edinburgh, Scotland, where the primary language is **English** (and Scots/Scottish Gaelic)."}]}
+{"id": "extras-016", "messages": [{"role": "user", "content": "Rate this answer.", "turn": 1}, {"role": "assistant", "content": "I'd rate it 4/5.", "weight": 1, "logprob": -0.21}]}
+{"id": "meta-017", "model": "lfm-8b", "temperature": 0.7, "reward": 0.83, "tags": ["coding", "python"], "messages": [{"role": "user", "content": "Hello"}, {"role": "assistant", "content": "Hi there! How can I help?"}]}
+[{"role": "user", "content": "Bare array conversation \u2014 no wrapper object."}, {"role": "assistant", "content": "Yep, this whole row is just a list of messages."}]
+{"id": "tooldefs-019", "messages": [{"role": "system", "content": "You can use these tools:\n[{\"type\": \"function\", \"function\": {\"name\": \"send_email\", \"description\": \"Send an email to a recipient.\", \"parameters\": {\"type\": \"object\", \"properties\": {\"to\": {\"type\": \"string\", \"description\": \"recipient email\"}, \"subject\": {\"type\": \"string\", \"description\": \"email subject\"}, \"body\": {\"type\": \"string\", \"description\": \"email body\"}}, \"required\": [\"to\", \"body\"]}}}]"}, {"role": "user", "content": "Email alice@example.com saying hi."}, {"role": "assistant", "content": "<|tool_call_start|>[send_email(to='alice@example.com', subject='Hi', body='Hello Alice!')]<|tool_call_end|>"}]}
+{"id": "markdown-020", "messages": [{"role": "user", "content": "Summarize the benefits of unit testing."}, {"role": "assistant", "content": "# Benefits of Unit Testing\n\nUnit testing offers several advantages:\n\n- **Catch bugs early** before they reach production\n- **Refactor safely** with a regression safety net\n- **Document behavior** \u2014 tests show how code is meant to be used\n\nUse `pytest` or `unittest` to get started. *Aim for fast, isolated tests.*"}]}

lql_cli-0.4.0/package-lock.json

@@ -0,0 +1,6 @@
+{
+  "name": "lql",
+  "lockfileVersion": 3,
+  "requires": true,
+  "packages": {}
+}

{lql_cli-0.2.1 → lql_cli-0.4.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "lql-cli"
-version = "0.2.1"
+version = "0.4.0"
 description = "lql — CLI for the Liquid DataViewer platform"
 readme = "README.md"
 requires-python = ">=3.9"
@@ -15,6 +15,7 @@ dependencies = [
     "httpx>=0.27",
     "rich>=13.0",
     "huggingface-hub>=0.24",
+    "textual>=0.50",
 ]
 
 [project.scripts]

lql_cli-0.4.0/src/lql/__init__.py

@@ -0,0 +1 @@
+__version__ = "0.4.0"

{lql_cli-0.2.1 → lql_cli-0.4.0}/src/lql/cli.py

@@ -15,7 +15,9 @@ from .commands import (
     highlights,
     instructions,
     issues,
+    preview,
     reports,
+    update as update_cmd,
     skills,
     spec,
     workspaces,
@@ -63,21 +65,12 @@ app.add_typer(reports.app, name="reports")
 app.add_typer(buckets.app, name="buckets")
 
 # Single commands
+app.command("preview", help="View dataset samples (local file or dataset ID) in the terminal")(preview.preview)
 app.command("instructions", help="Print a full reference for agents and humans")(instructions.instructions)
+app.command("update", help="Show how to update lql (or upgrade in place with --run)")(update_cmd.update)
 app.add_typer(skills.app, name="skills")
 
 
-@app.command("update")
-def update() -> None:
-    """Show how to update lql to the latest version."""
-    sys.stdout.write(
-        "To update lql, use the tool you installed it with:\n"
-        "  uv tool upgrade lql        # if installed via `uv tool install`\n"
-        "  pipx upgrade lql           # if installed via pipx\n"
-        "  pip install --upgrade lql  # if installed via pip\n"
-    )
-
-
 def main() -> None:
     # Mirror the TS top-level catch: unexpected errors surface as the documented
     # machine-readable contract, not a traceback. Typer/Click exits (SystemExit)

{lql_cli-0.2.1 → lql_cli-0.4.0}/src/lql/commands/instructions.py

@@ -71,6 +71,33 @@ A workspace is the top-level container for datasets, spec docs, and members.
 Upload a local file (uploads to HuggingFace liquid-ai org, then syncs):
   LQL_HF_TOKEN=<token> lql datasets upload <file.parquet> --workspace <id> --name <repo-name>
 
+## Preview
+
+View dataset samples in the terminal (a Textual TUI) — no browser. Chat-style
+layout (user right, system/assistant/tool left) with the same parsing as the
+DataViewer web UI: OpenAI {role, content}, structured/multimodal content
+([{type,text}] / image / audio segments), ShareGPT {from, value}, native OpenAI
+tool_calls, plus <think> reasoning blocks, <|tool_call_start|>…<|tool_call_end|>
+/ Python / XML / JSON tool calls, tool results, and tool-definition tables.
+Non-conversation fields render as labeled metadata.
+
+  lql preview <file.jsonl|file.json>     # local file: each line/object is a row
+  lql preview <dataset-id>               # platform dataset, fetched & paged lazily
+  lql preview <org/name> --hf            # HuggingFace repo: sync to DataViewer, then view
+  lql preview data.jsonl -c messages     # force field(s) as conversations
+
+With --hf, SOURCE is a HuggingFace repo (org/name): it's synced into a
+DataViewer workspace (you pick one interactively, or pass --workspace <id>;
+--split defaults to train) and reused on later previews (dedup by repo+split).
+
+Options: -c/--column (field(s) to treat as conversations; default auto-detect,
+repeatable), -n/--limit (page size when paging a platform dataset), --offset
+(start row index), --title, --hf, --split, --workspace, --profile, --api-url.
+
+Navigation: two modes toggled with m/Tab — pager (one sample at a time; ←/→ or
+n/b switch samples, ↑/↓/j/k scroll) and scroll (all samples; n/b jump between
+them). q quits. Works over plain SSH with no browser or port-forward.
+
 ## Evals
 
 Eval datasets (evaluation-run output: each row a sample with a model 'response'