TimesFM Forecasting

What it does

When your agent uses it

• Forecasting demand, sales or sensor readings without training a model
• Getting prediction intervals rather than a single forecast
• Adding holidays or promotions as covariates to a forecast
• Flagging unusual values in a time series

Example prompts

• “Forecast the next eight weeks of daily orders from orders.csv.”
• “Check whether my laptop can run TimesFM before we start.”
• “Predict energy demand with temperature and holidays as covariates.”

Requirements

• Python with the timesfm package
• Enough RAM for the model (the checker blocks below 2 GB)

Workflow steps

3 steps, taken from the step headings in SKILL.md.

1. 1Verify System (always first)
2. 2Install TimesFM
3. 3Install PyTorch for Your Hardware

What it can do on your machine

Read from SKILL.md and the folder at commit e51928e. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Ships 1 file in scripts/ (Python, from the files we listed), which the agent can run.

  Shell commands in SKILL.md call:

  • python
  • pip
  • uv

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  Hosts in commands or code, which the agent is likely to contact:

  • download.pytorch.org

  Also links to:

  • arxiv.org
  • huggingface.co
  • research.google
  • cloud.google.com

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

TimesFM Forecasting loads about 4.7k tokens when it runs, and up to ~11k if it reads all its reference files. Until then it costs about 193 tokens; SKILL.md has 1,170 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~193

When it runs · the whole SKILL.md, loaded when a task matches

~4.7k

With references · SKILL.md plus every file in references/, read only if the agent opens them

~11k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from google-research/timesfm at commit e51928e, republished under its Apache-2.0 licence (© google-research). 1,170 words, ~4,680 tokens.

name

timesfm-forecasting

description

Zero-shot time series forecasting with Google's TimesFM foundation model. Use this skill when forecasting ANY univariate time series — sales, sensor readings, stock prices, energy demand, patient vitals, weather, or scientific measurements — without training a custom model. Supports both basic forecasting and advanced covariate forecasting (XReg) with dynamic and static exogenous variables. Automatically checks system RAM/GPU before loading the model, validates dataset fit before processing, supports CSV/DataFrame/array inputs, and returns point forecasts with calibrated prediction intervals. Includes a preflight system checker script that MUST be run before first use to verify the machine can load the model and handle your specific dataset.

license

Apache-2.0

metadata.author

Clayton Young (@borealBytes)

metadata.version

1.0.0

python scripts/check_system.py

flowchart TD
    start["🚀 Run check_system.py"] --> ram{"RAM ≥ 4 GB?"}
    ram -->|"Yes"| gpu{"GPU available?"}
    ram -->|"No (2-4 GB)"| warn_ram["⚠️ Warning: tight RAM<br/>CPU-only, small batches"]
    ram -->|"No (< 2 GB)"| block["🛑 BLOCKED<br/>Insufficient memory"]
    warn_ram --> disk
    gpu -->|"CUDA / MPS"| vram{"VRAM ≥ 2 GB?"}
    gpu -->|"CPU only"| cpu_ok["✅ CPU mode<br/>Slower but works"]
    vram -->|"Yes"| gpu_ok["✅ GPU mode<br/>Fast inference"]
    vram -->|"No"| cpu_ok
    gpu_ok --> disk{"Disk ≥ 2 GB free?"}
    cpu_ok --> disk
    disk -->|"Yes"| ready["✅ READY<br/>Safe to load model"]
    disk -->|"No"| block_disk["🛑 BLOCKED<br/>Need space for weights"]

# Quick estimate for your dataset
python scripts/check_system.py \
  --num-series 1000 \
  --context-length 1024 \
  --horizon 24 \
  --batch-size 32 \
  --estimate-only

Total CPU memory: 2.34 GB
Total GPU memory: 2.15 GB

Dataset requires ~12.5 GB RAM but system has 8.0 GB.
Try: context_length=512 or process in chunks of 50 series.

python scripts/check_system.py

# Using uv (fast)
uv pip install timesfm[torch]

# Or using pip
pip install timesfm[torch]

# For JAX/Flax backend (faster on TPU/GPU)
uv pip install timesfm[flax]

# CUDA 12.1 (NVIDIA GPU)
pip install torch>=2.0.0 --index-url https://download.pytorch.org/whl/cu121

# CPU only
pip install torch>=2.0.0 --index-url https://download.pytorch.org/whl/cpu

# Apple Silicon (MPS)
pip install torch>=2.0.0  # MPS support is built-in

import torch, numpy as np, timesfm

torch.set_float32_matmul_precision("high")

model = timesfm.TimesFM_2p5_200M_torch.from_pretrained(
    "google/timesfm-2.5-200m-pytorch"
)
model.compile(timesfm.ForecastConfig(
    max_context=1024, max_horizon=256, normalize_inputs=True,
    use_continuous_quantile_head=True, force_flip_invariance=True,
    infer_is_positive=True, fix_quantile_crossing=True,
))

point, quantiles = model.forecast(horizon=24, inputs=[
    np.sin(np.linspace(0, 20, 200)),  # any 1-D array
])
# point.shape == (1, 24)         — median forecast
# quantiles.shape == (1, 24, 10) — 10th–90th percentile bands

point, quantiles = model.forecast_with_covariates(
    inputs=inputs,
    dynamic_numerical_covariates={"price": price_arrays},
    dynamic_categorical_covariates={"holiday": holiday_arrays},
    static_categorical_covariates={"region": region_labels},
    xreg_mode="xreg + timesfm",  # or "timesfm + xreg"
)

point, q = model.forecast(horizon=H, inputs=[values])

lower_90 = q[0, :, 1]  # 10th percentile
upper_90 = q[0, :, 9]  # 90th percentile

actual = test_values
anomalies = (actual < lower_90) | (actual > upper_90)

point, q = model.forecast(horizon=H, inputs=data)

lower_80 = q[:, :, 1]  # 10th percentile
upper_80 = q[:, :, 9]  # 90th percentile
median   = q[:, :, 5]

timesfm.ForecastConfig(
    max_context=1024,                    # Max context window
    max_horizon=256,                     # Max forecast horizon
    normalize_inputs=True,               # RECOMMENDED — prevents scale instability
    per_core_batch_size=32,              # Tune for memory
    use_continuous_quantile_head=True,   # Better quantile accuracy for long horizons
    force_flip_invariance=True,          # Ensures f(-x) = -f(x)
    infer_is_positive=True,              # Clamp forecasts ≥ 0 when all inputs > 0
    fix_quantile_crossing=True,          # Ensure q10 ≤ q20 ≤ ... ≤ q90
    return_backcast=False,               # Return backcast (for covariate workflows)
)

import torch, numpy as np, pandas as pd, timesfm, matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt

torch.set_float32_matmul_precision("high")
model = timesfm.TimesFM_2p5_200M_torch.from_pretrained(
    "google/timesfm-2.5-200m-pytorch"
)
model.compile(timesfm.ForecastConfig(
    max_context=512, max_horizon=52, normalize_inputs=True,
    use_continuous_quantile_head=True, fix_quantile_crossing=True,
))

df = pd.read_csv("weekly_demand.csv", parse_dates=["week"])
values = df["demand"].values.astype(np.float32)

point, quantiles = model.forecast(horizon=52, inputs=[values])

fig, ax = plt.subplots(figsize=(12, 5))
ax.plot(values[-104:], label="Historical")
x_fc = range(len(values[-104:]), len(values[-104:]) + 52)
ax.plot(x_fc, point[0], label="Forecast", color="tab:orange")
ax.fill_between(x_fc, quantiles[0, :, 1], quantiles[0, :, 9],
                alpha=0.2, color="tab:orange", label="80% PI")
ax.legend(); ax.set_title("52-Week Demand Forecast")
plt.tight_layout(); plt.savefig("forecast.png", dpi=150)

df = pd.read_csv("all_stores.csv", parse_dates=["date"], index_col="date")
inputs = [df[col].dropna().values.astype(np.float32) for col in df.columns]

point, quantiles = model.forecast(horizon=30, inputs=inputs)

import json
results = {col: {"forecast": point[i].tolist(),
                 "lower_80": quantiles[i, :, 1].tolist(),
                 "upper_80": quantiles[i, :, 9].tolist()}
           for i, col in enumerate(df.columns)}
with open("batch_forecasts.json", "w") as f:
    json.dump(results, f, indent=2)

H = 24
train, actual = values[:-H], values[-H:]
point, quantiles = model.forecast(horizon=H, inputs=[train])
pred = point[0]

mae  = np.mean(np.abs(actual - pred))
rmse = np.sqrt(np.mean((actual - pred) ** 2))
mape = np.mean(np.abs((actual - pred) / actual)) * 100
coverage = np.mean((actual >= quantiles[0, :, 1]) & (actual <= quantiles[0, :, 9])) * 100

print(f"MAE: {mae:.2f} | RMSE: {rmse:.2f} | MAPE: {mape:.1f}% | 80% PI Coverage: {coverage:.1f}%")

# Always set on Ampere+ GPUs (A100, RTX 3090+)
torch.set_float32_matmul_precision("high")

# Batch size guidelines:
# GPU 8 GB VRAM:  per_core_batch_size=64
# GPU 16 GB VRAM: per_core_batch_size=128
# CPU 8 GB RAM:   per_core_batch_size=8
# CPU 16 GB RAM:  per_core_batch_size=32

# Memory-constrained: process in chunks
CHUNK = 50
results = []
for i in range(0, len(inputs), CHUNK):
    p, q = model.forecast(horizon=H, inputs=inputs[i:i+CHUNK])
    results.append((p, q))

# Basic system check
python scripts/check_system.py

# Check if your specific dataset will fit
python scripts/check_system.py \
  --num-series 1000 \
  --context-length 1024 \
  --horizon 24 \
  --batch-size 32

# Quick memory estimate without system checks
python scripts/check_system.py \
  --num-series 5000 \
  --context-length 2048 \
  --estimate-only

python scripts/forecast_csv.py input.csv \
    --horizon 24 \
    --date-col date \
    --value-cols sales,revenue \
    --output forecasts.csv

# Run all three examples:
cd examples/global-temperature && python run_forecast.py && python visualize_forecast.py
cd examples/anomaly-detection  && python detect_anomalies.py
cd examples/covariates-forecasting && python demo_covariates.py

# Anomaly detection regression:
python -c "
import json
d = json.load(open('examples/anomaly-detection/output/anomaly_detection.json'))
assert d['context_summary']['critical'] >= 1, 'Sep 2023 must be CRITICAL'
print('Anomaly detection: PASS')"

# Covariates regression:
python -c "
import pandas as pd
df = pd.read_csv('examples/covariates-forecasting/output/sales_with_covariates.csv')
assert len(df) == 108, f'Expected 108 rows, got {len(df)}'
print('Covariates: PASS')"

© google-research, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 31 other files (scripts, references) in timesfm-forecasting of google-research/timesfm.

• SKILL.md
• examples/anomaly-detection/detect_anomalies.py
• examples/anomaly-detection/output/anomaly_detection.json
• examples/anomaly-detection/output/anomaly_detection.png
• examples/covariates-forecasting/demo_covariates.py
• examples/covariates-forecasting/output/covariates_data.png
• examples/covariates-forecasting/output/covariates_metadata.json
• examples/covariates-forecasting/output/sales_with_covariates.csv
• examples/finetuning/README.md
• examples/finetuning/finetune_lora.py
• examples/global-temperature/README.md
• examples/global-temperature/generate_animation_data.py
• examples/global-temperature/generate_gif.py
• examples/global-temperature/generate_html.py
• … and 18 more

Open the folder on GitHubat commit e51928e